Author Topic: Re: eht for ion Rb (Read 9756 times)

yuniawan · « **on:** October 29, 2015, 05:55:53 AM »

dear. I always filed making structure of RbNH3, i will simulate ion Rb in NH3 liquid.
i use this coordinates
$coord
0.00000000000000 0.00000000000000 4.00000000000000 rb
0.00000000000000 0.00000000000000 0.00000000000000 n
0.00000000000000 1.77067701000000 -0.72376659000000 h
1.53446076000000 -0.88439364000000 -0.72376659000000 h
-1.53446076000000 -0.88439364000000 -0.72376659000000 h
$user-defined bonds

but always file in EHT process, the message is like here

---------------
...
...
reading orbital data 4S(DZ) from file /usr/local/tm/basen/n .

PGFIO-F-231/formatted read/unit=9/error on data conversion.
File name = /usr/local/tm/basen/n formatted, sequential access record = 1058
In source file rdmo.f, at line number 276
-------------------

basisset of NH3 should be DZP or DZP (dunning) becouse we have match the result with experimental data
I have changed basisset of Rb with several types but always stuck with the same message above.

below is my define
-------------------

rbnh3
sy c1
a coord
*
no
b "n" "h" DZP(Dunning)
file basisdzpd.tbs
file basisdzpd.tbs
b "rb" Def2-SVP
file basisdefsvp.tbs
ecp "rb" Def2-ECP
file basisdefsvp.tbs
*
eht
y
y
1
y
prop

6

y
netto
*

*
scf
thi
99999
99999
iter
3000

drv
-sec -dip -pol
*
*
*
*
-------------

can you give any idea? thanks. ( It have been three month I faced this problem)

uwe · « **Reply #1 on:** October 29, 2015, 12:33:24 PM »

Hello,

where do you run define on? Is it a Windows system or a Linux machine? The error seems to stem from Windows, but the path is Unix style...

Could be a problem with DOS/Unix formatted ASCII files - so if you run in under Linux but have the basis set files from Windows, you should first run dos2unix in /usr/local/tm/basen/ for all files there.

Regards,

Uwe

yuniawan · « **Reply #2 on:** October 29, 2015, 03:38:20 PM »

i run under linux. When i use different ion such potassium with similar coordinat everything goes normaly.

uwe · « **Reply #3 on:** October 29, 2015, 04:20:49 PM »

Hello,

if DZP(Dunning) is the name of the basis set as you get it from the EMSL basis set library https://bse.pnl.gov/bse/portal, then I wonder how this one looks like for Rb. Could you post the basis set here? Does it include ECPs?

Regards,

Uwe

yuniawan · « **Reply #4 on:** October 29, 2015, 10:36:41 PM »

For Rb i use def2-svp or lanl2dz ecp basis set. I take it from emsl also. DZP dunning only for nh3.

uwe · « **Reply #5 on:** October 30, 2015, 09:52:19 AM »

Hi,

define is not able to read in the file, so there seems to be something wrong with it. Please do post the file here. We will then probably be able to reproduce the error.

Regards,

Uwe

yuniawan · « **Reply #6 on:** October 30, 2015, 04:04:40 PM »

this is a file coord consit of RbNH3 coordinates

$coord
-0.00000900000000 -0.67251400000000 0.00000000000000 rb
-0.00000900000000 2.38987600000000 0.00000000000000 n
0.94928000000000 2.71735600000000 0.00000000000000 h
-0.47445000000000 2.71825600000000 0.82183500000000 h
-0.47445000000000 2.71825600000000 -0.82183500000000 h
$user-defined bonds
$end

below is content of my define, I used lanl2dz ecp for Rb and DZP for N and H

rbnh3
sy c1
a coord
*
no
b "n" "h" DZP
file basis.tbs
b "rb" LANL2DZ ECP
file basis.tbs
ecp "rb" LANL2DZ ECP
file basis.tbs
*
eht
y
y
1
y
prop

6

y
netto
*

*
scf
thi
99999
99999
iter
3000

drv
-sec -dip -pol
*
*
*
*

i make a file which name is basis.tbs and it contain basis set of RbNH3

$basis
*
h DZP
*
3 s
12.98511484 0.0310449
1.95908134 0.2173262
0.44403649 0.7516287
1 s
0.11282695 1.0000000
1 p
0.80305095 1.0000000
*
n DZP
*
5 s
6711.76398036 0.0019838
1029.56088507 0.0148619
234.62552896 0.0761291
65.08616322 0.2866453
20.68134371 0.7108440
2 s
7.35948620 0.7617357
2.86133058 0.2682842
1 s
0.75772555 1.0000000
1 s
0.22278085 1.0000000
4 p
26.95312900 0.0185616
6.01777874 0.1165761
1.76062640 0.3816428
0.56064983 0.6412379
1 p
0.17525619 1.0000000
1 d
0.89591019 1.0000000
*
rb LANL2DZ ECP
*
3 s
1.4570000 -1.0617406
0.9400000 1.1722125
0.2706000 0.7437008
4 s
1.4570000 0.5052174
0.9400000 -0.5718404
0.2706000 -0.6835308
0.0366000 1.1136477
1 s
0.0155000 1.0000000
3 p
3.3000000 -0.0729417
0.5985000 0.6321782
0.2067000 0.4707426
2 p
0.1947000 -0.1250552
0.0318000 1.0438206
1 p
0.0124000 1.0000000
*

$ecp
*
rb LANL2DZ ECP
*
ncore = 28 lmax = 3
f-ul
-0.0364362 0 879.5983664
-20.8033052 1 142.0188695
-115.4661910 2 40.9346882
-42.5061293 2 9.9084802
-5.5247889 2 3.1556041
s-ul
2.9747910 0 70.9807215
36.1170435 1 38.6949807
152.2741068 2 15.9082035
53.7151906 2 3.7152659
p-ul
4.9551026 0 93.3058684
24.2726236 1 47.1221370
215.2341411 2 48.8050640
151.6687666 2 14.3784949
33.8750280 2 3.0547807
d-ul
3.0465224 0 33.0477430
29.4243624 1 9.9601880
33.4431808 2 3.7403577
11.1029460 2 0.8124208
*
$end

if you change the ion Rb with others (ex. K or Cs) with same coordinates using any basisset then we will not find this errors.

uwe · « **Reply #7 on:** October 30, 2015, 05:09:59 PM »

Hi,

the older version I was able to use (without digging in backups) was Turbomole 5.10

With the define version of version 5.10 the input was done without errors using your basis set file. An energy calculation also worked, see attached output.

I'd recommend to use a new Turbomole version. define from the latest version can be used if you download the TmoleX client version which includes everything that is necessary to generate input files:

http://www.cosmologic.de/tmolex-client

Regards

Uwe

yuniawan · « **Reply #8 on:** November 02, 2015, 02:24:45 PM »

thanks a lot, it seem the problem is solved as your advice.