CCL: TM5.10: problems with constrained geometry optimization during a search for a transition state



 Sent to CCL by: Aleksey Kuznetsov [AlexKuznetsov2007,,yandex.ru]
 Dear CCL Users,
 I am looking for a transition state for the reaction of H2 with a N2(transition
 metal cluster) adsorbed at MgO(100) surface, using the embedded cluster approach
 (MgO cluster is embedded into the point charge array) with Turbomole 5.10. In my
 calculation, the Mg's and O's atoms are fixed, and only H2, N2, and transition
 metal cluster are allowed to relax (one distance N-H is also fixed). Below there
 is a part of my 'coord' file:
 $coord
     2.39032877604133      8.25913300313270     -7.32768851332685  h
     3.33127069298153      9.83887033461485     -5.58504005522075  h
    -0.17492593800043      5.52925592375161     -7.10524544210555  n
    -0.04705169228237      8.16658554084386     -8.08372916213718  n
     0.02396047105540      2.36709809807531     -3.74432586066866  zr
    -2.76503557449325      6.07384486515057     -1.92546840309758  pd
     1.97586710418695      6.90664433702049     -4.17446765970979  zr
     4.75623866501232      2.85042595434685     -2.04843366515992  pd
    -1.64927570145236      0.65155657107799      6.53597754772781  mg     f
    -1.04006469510789      3.40790675809863      3.73058912774775  o      f
    -3.94370461903670     -1.40503558837978      4.01708230923380  o      f
    -3.33449361269223      1.35131459864087      1.21169388925374  mg     f
    -6.23813353662104     -3.46162774783755      1.49818707073978  mg     f
    -5.62892253027656     -0.70527756081690     -1.30720134924028  o      f
     1.54497952983392     -1.35138399890064      5.26170474285745  o      f
     2.15419053617839      1.40496618812000      2.45631632287739  mg     f
    -0.74944938775042     -3.40797615835841      2.74280950436344  mg     f
    -0.14023838140595     -0.65162597133776     -0.06257891561662  o      f
    -3.04387830533476     -5.46456831781618      0.22391426586942  o      f
    -2.43466729899029     -2.70821813079553     -2.58147415411064  mg     f
     4.73923476112019     -3.35432456887927      3.98743193798709  mg     f
     5.34844576746467     -0.59797438185863      1.18204351800703  o      f
     2.44480584353585     -5.41091672833704      1.46853669949308  o      f
     3.05401684988033     -2.65456654131639     -1.33685172048699  mg     f
     0.15037692595151     -7.46750888779481     -1.05035853900094  mg     f
     0.75958793229598     -4.71115870077416     -3.85574695898100  o      f
     6.24827192949393     -4.65750701618949     -2.61112446485111  o      f
     5.63906092314945     -7.41385720321014      0.19426395512895  mg     f
    -6.52874869230583      3.35425507351396      2.48596663361786  o      f
    -7.13795969865030      0.59790488649332      5.29135505359792  mg     f
    -4.72909610762837     -4.76481019260023     -5.10036927300007  o      f
     4.44861934481647      3.46155824992470      4.97521144176682  o      f
    -5.33830711397284     -7.52116037962087     -2.29498085302001  mg     f
     3.83940833847200      0.70520806290406      7.78059986174688  mg     f
     9.44252716078020     -6.66044758616812     -3.88539726972147  mg     f
     3.95384301190959     -6.71409917564726     -5.13001970334513  mg     f
    -1.53484087634209     -6.76775076257886     -6.37464207787043  mg     f
    -7.02352502521271     -6.82140235205799     -7.61926451149408  mg     f
    -9.72300392359210      5.35719564349259      3.76023943848822  mg     f
    -8.82317760989017      1.29766291405620     -0.03292860487615  mg     f
    -7.92335133891464     -2.76186962262160     -3.82609646812971  mg     f
    -4.23431977472149      5.41084723297173      5.00486187211187  mg     f
     1.25436411353019      5.46449881990333      6.24948424663718  mg     f
     6.74304826240081      5.51815040938247      7.49410668026083  mg     f
     7.64287457610274      1.45861767994607      3.70093863689646  mg     f
     8.54270084707827     -2.60091485673172     -0.09222922635710  mg     f
     1.83559486611852     -8.16726678447804      4.27392498626501  mg     f
    -4.55291559645403     -4.16138564452078      6.82247059600573  mg     f
    -3.65308928275210     -8.22091837395717      3.02930255264136  mg     f
     0.93576855241658     -4.10773405504164      8.06709302962939  mg     f
 $intdef
 # definitions of internal coordinates
    1 f  1.0000000000000 stre    1    4           val=   2.55362
 $user-defined bonds
 $redundant
      number_of_atoms            50
      degrees_of_freedom        144
      internal_coordinates      323
      frozen_coordinates          1
 # definitions of redundant internals
 . . .
 I use RI-B3LYP method with TZVP basis set (and ECP's for Zr's and Pd's). For
 constained geometry optimization, I am using the command "jobex -ri -c 500
 -statpt".
 However, I am quite confused by the output results. First of all, at the some
 step of optimization, the energy of the system becomes lower than the sum of
 energies of H2 molecules and original N2Zr2Pd2/MgO system, and the gradient of
 the system becomes huge:
  energy change  : actual value =   -16.27     threshold =   0.1000E-05
  geom. gradient : actual value =    358.9     threshold =   0.1000E-02
 Could anybody please explain why such strange huge values could appear, and
 which which part of the output should be used for reference? Should I look at
 this part of the 'job.last' file:
       ******************************************************************
                           CONVERGENCE INFORMATION
                                Converged?     Value      Criterion
              Energy change         no      16.2682530   0.0000010
              RMS of displacement   no       0.0250000   0.0005000
              RMS of gradient       no      42.3591183   0.0005000
              MAX displacement      no       0.1769108   0.0010000
              MAX gradient          no     358.9048907   0.0010000
       ******************************************************************
 Is there any other way how to do search for a transition state in such cases -
 for on-surface reactions?
 Thank you very much in advance.
 With best regards,
 Aleksey Kuznetsov.
 -------------------------------------------
 Dr. Aleksey Kuznetsov
 Cherry L. Emerson Center for Scientific Computation
 Emory University
 1515 Dickey Drive
 Atlanta, GA 30322
 USA
 Phone: (404)727-2381
 Email: akuznets[A]euch4e.chem.emory.edu
 -------------------------------------------