ZMAT and Symmetry

Find answers or ask questions regarding Dalton calculations.
Please upload an output file showing the problem, if applicable.
(It is not necessary to upload input files, they can be found in the output file.)

Post Reply
cymantren
Posts: 10
Joined: 11 Jan 2014, 13:59
First name(s): Alfred
Last name(s): Güthler
Affiliation: privat
Country: Germany

ZMAT and Symmetry

Post by cymantren » 08 Jul 2019, 08:06

Hello DALTON-users,

I have problems using symmetry AND mol-files, especially with the ZMAT- format, which I prefer. Using a simple 3-atomic molecule, with C2v-sym. , the program recognizes it as one with 4 atoms ? ! How do I circumvent that, so that the correct molecule is used ?
If I use cartesian coord. I must have to use less lines in the mol file, because of the symmetry. That means not three, but 2 ? How do I know, which line I have to drop ?

Thank you in advance for your help and answer.

best greetings

Alfred
Attachments
optvib_B3_631gdp_zmt-C2v.out
(50.57 KiB) Downloaded 17 times

User avatar
magnus
Posts: 462
Joined: 27 Jun 2013, 16:32
First name(s): Jógvan Magnus
Middle name(s): Haugaard
Last name(s): Olsen
Affiliation: Hylleraas Centre, UiT The Arctic University of Norway
Country: Norway

Re: ZMAT and Symmetry

Post by magnus » 08 Jul 2019, 08:11

It's because you're using symmetry generators which will apply the specified symmetry operations. Instead remove the generators and let Dalton detect the symmetry.

cymantren
Posts: 10
Joined: 11 Jan 2014, 13:59
First name(s): Alfred
Last name(s): Güthler
Affiliation: privat
Country: Germany

Re: ZMAT and Symmetry

Post by cymantren » 22 Jul 2019, 19:56

Dear Magnus,

thank you for your reply !
Using your suggestion, the calculation works very well, and the symmetry D3h --> C2v is correctly detected by DALTON2018.
I know however, that I have to use symmetry adapted coordinates, in connection with symmetry, at least with cartesians, because of the symmetry generators, but I wanted to now, how this work with zmatrices ?!
Do you have any idea and is there any way to to that ?! It is sometimes useful to calculate a molecule at a certain symmetry. Using tables that shows, how the symmetry is reduced - D3h --> C2v - is also a possibility.
In the case of B3, the correct symmetry is D3h.


Thank you in advance for your answer.

best greetings

Alfred Güthler
Attachments
optvib_B3_631gstst_zmat.out
(142.63 KiB) Downloaded 14 times

kennethruud
Posts: 252
Joined: 27 Aug 2013, 16:42
First name(s): Kenneth
Last name(s): Ruud
Affiliation: UiT The Arctic University of Norway
Country: Norway

Re: ZMAT and Symmetry

Post by kennethruud » 23 Jul 2019, 12:14

Hi!

I am a little bit uncertain what your actual question is, but in general Dalton provides little support for the use of Z-matrices, and they are mainly provided for those that really do not want to work with Cartesian coordinates.

The internal handling of symmetry in Dalton is all based on simple bit operations for handling symmetry, which basically means we handle D2h and subgroups and only symmetry operations that change the sign of one or more of the Cartesian coordinates of a given atom and nothing more (programs able to handle non-Abelian symmetry often use petite lists).

Thus, you can very easily define symmetry generators in your input (e.g. Z, indicating that we will generate a symmetry-related atom for any atom for which the Z coordinate =/= 0, that is, a mirror plane); or XY, indicating we will change simultaneously change the sign of the X and Y coordinates of atoms where one or both of these coordinates =/=0, that is, a rotation about the Z axis).

However, all this requires that the Cartesian axis system also constitute the rotation axes and mirror planes and that the molecule is oriented symmetrically with respect to these symmetry operations. That is also why there is no straightforward way of coupling a Z matrix input to a specific symmetry input. Instead, what the program does, as you have seen, is to move the molecular centre of mass to the origin of a Cartesian axes system, and align the moments of inertia tensor components along the different Cartesian coordinate axis as a way of detecting when symmetry is after all present.

As I said, I am uncertain whether this answers your question, but nevertheless hopefully of some help.


Best regards,

Kenneth

cymantren
Posts: 10
Joined: 11 Jan 2014, 13:59
First name(s): Alfred
Last name(s): Güthler
Affiliation: privat
Country: Germany

Re: ZMAT and Symmetry

Post by cymantren » 24 Jul 2019, 18:42

Dear Kenneth,

I have read your answer and I can understand that you are a bit uncertain, about my question, because I was not very clear at some points. However, as you hoped, you partially gave me an answer.
My Question was/is: how can one use a ZMATRIX-input in a mol-file at a certain symmetry, if one already knows or guess the correct symmetry. ?
For instance, using H20 with the zmatrix at C2v-symmetry, one can be certain that the calculation is correct, because H20 HAS C2v-sym.
Actually I could do any calculation without symmetry-generators, but what would happen, if I use these "Generators = 2 X Z" ?
I will try this with H2CO.

Best Greetings

Alfred

taylor
Posts: 545
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Tianjin University
Country: China

Re: ZMAT and Symmetry

Post by taylor » 25 Jul 2019, 02:44

It is inherent in the definitions of the terms in a Z-matrix that dealing with symmetry is complicated. It usually requires introduction of dummy atoms to ensure things like proper ring closure for symmetric rings like benzene, or with linear sections of a molecule. Dalton will detect the (highest subgroup of D2h) symmetry from a Z-matrix, as other posters have said. I do not understand why you want to use generators, because they require your input to comprise only symmetry-unique atoms.

So for H2CO it would be possible to use a Z-matrix (and in this case no dummy atoms would be required), and if the geometry specified in the Z-matrix has C2v symmetry (and the input does not contain the Nosymmetry keyword) the program will detect this symmetry and use it. However, if you tell the program to use two generators to define C2v, then the code expects only three atoms in the input because it will generate the second H atom. I have no idea how to do this with a Z-matrix, and to be honest I can't understand why one would want to.

I admit to very considerable prejudice here. I loathe Z-matrices and regard them as the worst thing Jimmy Stewart has contributed to our field. They were targetted at systems that have no symmetry, and/or for use in codes that do not use symmetry. I recognize that the most commonly used ab initio program uses Z-matrices and has popularized their use, but just because a majority of people do it does not make it sensible (a majority of people who voted supported the UK leaving the EU: this provides another illustration...). I suggest that if you want to exploit symmetry --- and it is almost always advantageous to do so --- you use Cartesian coordinates and specify the symmetry yourself with generators. The posting that appears at the top of the "Running Dalton" topic provides some description of this (full disclosure: I wrote it...).

Best regards
Pete

cymantren
Posts: 10
Joined: 11 Jan 2014, 13:59
First name(s): Alfred
Last name(s): Güthler
Affiliation: privat
Country: Germany

Re: ZMAT and Symmetry

Post by cymantren » 31 Jul 2019, 12:36

Dear Users,

The reason why I use zmatrices is the easy us e of it and its conection with the internal coordinates. For Instance in the case of H20 one has: 2A_1+B_1, two different r's and one angle. That is exactly how one (can) write the z-matrix for water.
There is another reason: I can imagine the molecule in z-matrix-format at all, and not in cartesians. One had to "translate" the cartesian to bondings and angles, to get an imagination, about what happens. This is not possible without the help of other programs.

Nevertheless, thank you all for you valuable help.

Best Greetings

Alfred

cymantren
Posts: 10
Joined: 11 Jan 2014, 13:59
First name(s): Alfred
Last name(s): Güthler
Affiliation: privat
Country: Germany

Re: ZMAT and Symmetry

Post by cymantren » 31 Jul 2019, 12:51

Hello users,

I have to correct, what I have said before, concerning the vibrations.
The case H2O:
2A1 + B1.
There are two A1, namly the symmetric bonding, angle and one antisymmetric bonding, which is of B1.sym.
In the case B3:
D3h: A'_1+E' which is reduced in DALTON to C2v-sym.


greetings
Alfred

hjaaj
Posts: 317
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: ZMAT and Symmetry

Post by hjaaj » 01 Aug 2019, 21:27

Dear Alfred

To summarize:
1) When using ZMAT you must specify all atoms in the Z matrix and you must let Dalton determine the symmetry, i.e. it is not allowed to specify any symmetry generators. It is allowed to specify NOSYMMETRY. Any dummy atoms are eliminated, so that is not a problem.
2) You can see the definition of the symmetry-adapted coordinates in the output.

cymantren
Posts: 10
Joined: 11 Jan 2014, 13:59
First name(s): Alfred
Last name(s): Güthler
Affiliation: privat
Country: Germany

Re: ZMAT and Symmetry

Post by cymantren » 03 Aug 2019, 08:10

Hello Mr. Jensen and those who have written to me,

your answer is a bit pitty, however it (seems) that it is simply not possible in Dalton to use generators in connection with a zmatrix, not even with dummy atoms ?!
I have already written, why I am "fond of" zmatrices, although cartesians are prefered.
However your answer is the final one, so I have to look for an alternative.

Thank you and all !

Greetings

Alfred

hjaaj
Posts: 317
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: ZMAT and Symmetry

Post by hjaaj » 04 Aug 2019, 20:35

Alfred, but you specify all atoms in a z-matrix, right? As in your example. And as you can see in the second output you uploaded, when you let Dalton detect symmetry, it reduces automatically the total number of atoms to the number of symmetry-independent atoms and generates the remaining atoms with the found symmetry operations. In your example the third B-atom is eliminated

Code: Select all

 Symmetry Independent Centres             
 ----------------------------
B  2   5 :     -0.98193050     1.70075351     0.00000000  Isotope  1
B  1   5 :      1.96386100     0.00000000     0.00000000  Isotope  1
and later restored by the symmetry operations (B 2/ 1 and B 2/ 2):

Code: Select all

  Cartesian Coordinates (a.u.)
  ----------------------------

  Total number of coordinates:    9
  B  1    :     1  x   1.9638609965    2  y   0.0000000000    3  z   0.0000000000
  B  2/ 1 :     4  x  -0.9819304983    5  y   1.7007535125    6  z   0.0000000000
  B  2/ 2 :     7  x  -0.9819304983    8  y  -1.7007535125    9  z   0.0000000000
Dalton also tells you the symmetry-adapted nuclear coordinates it is using in the C_2v subgroup (the coordinate numbers refers to the numbers in the list of Cartesian coordinates above):

Code: Select all

  Symmetry Coordinates
  --------------------

  Number of coordinates in each symmetry:     3    3    2    1

  Symmetry  A1  ( 1)

    1   B  1  x    1
    2   B  2  x    [  4  +    7 ]/2
    3   B  2  y    [  5  -    8 ]/2

  Symmetry  B1  ( 2)

    4   B  1  y    2
    5   B  2  x    [  4  -    7 ]/2
    6   B  2  y    [  5  +    8 ]/2

  Symmetry  B2  ( 3)

    7   B  1  z    3
    8   B  2  z    [  6  +    9 ]/2

  Symmetry  A2  ( 4)

    9   B  2  z    [  6  -    9 ]/2
I do not understand what else you want, but if this is not what you want, good luck with finding an alternative.

cymantren
Posts: 10
Joined: 11 Jan 2014, 13:59
First name(s): Alfred
Last name(s): Güthler
Affiliation: privat
Country: Germany

Re: ZMAT and Symmetry

Post by cymantren » 08 Aug 2019, 12:07

Dear Mr. Hans Jorgen Aagaard Jensen and other USERS,

you are definitely right. I simply have overlooked this fact, that the symmetry coordinates al already present and the unique's also. What I really wanted was, to calculate a molecule - B3 - at a certain symmetry and let not Dalton find out the symmetry. That's all. At least I see and see now, that the alternative I was looking for, lies before me.
It would be very kind, if you could tell me, the way I can use the symmetry coordinates to generate the symmetryadapted equations - normal coordinates..
What I know is, that there should be 3 nomal-coordinates of A1 and B1 sym. The irreducible reps are: 2A1+B1, at least for an AB2 molecule.
Moreover if Ia = Ib < Ic, means that it is an oblate rotator. However the calc. at symmetry C2v, hints on an asymetric rotator, because of the moment of inertia at line 3483 which is: Ia < Ib < Ic ?!
Is this an effect of the choosen symmetry - OR ?
Actually I wanted to use D3h, which should be an oblate rotator, but the calculation uses C2v and gives an asymetric rotator. So I 'm a little bit confused about the results, mainly about the different kind of rotators
I am wondering also about the energy of the MOs. There should be also some degenerate orbs (---> lines 2908). There are some ALMOST degenerate orbs, which should give E'-orbs - but only almost. I expected some MOs of different sym. but with the same energy value. For Instance, E' splits into A1+B2 and A1+B2 of the same energy, must be an E' orb ?

My questions is, how does the reduced symmetry influences the outcome of the calculation ?!

Please give me some hints and tips. Thank you !

Best greetings

Alfred

taylor
Posts: 545
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Tianjin University
Country: China

Re: ZMAT and Symmetry

Post by taylor » 08 Aug 2019, 13:15

Can I start by suggesting you use A1 and B2 for the "in-plane" irreducible representations? This is consistent with Mulliken's attempt to standardize these things in 1955. It is essential not to blindly accept the labels the program prints out in the C2v and D2h cases. These are not uniquely definable by the program and are in effect best guesses that are sometimes not consistent with Mulliken's proposal. (See the manual and my forum posting that I referred to before.)

Now, you are absolutely correct that the geometry is wandering away from D3h to C2v, and that the major cause of this relates to your starting with a geometry that would have a partly occupied degenerate orbital. Doing an SCF calculation here is problematic because the program will make a decision about which component of the degenerate orbital it occupies. This is then no longer an irreducible symmetry state (in D3h) and the calculation can only preserve the symmetry for this state, which is C2v. As you see you converge to a state that has a dipole moment.

One way to ensure that you maintain three-fold symmetry is to do an MCSCF calculation using the SUPSYM (supersymmetry) option. When the code recognizes that a higher symmetry is present it will attempt to retain that symmetry, although sometimes this can require tightening the (already demanding) default convergence thresholds.

If you are simply experimenting, a possibility might be to do, say, B3+, which I think would be a closed-shell system. In this case an SCF geometry optimization should preserve the D3h symmetry. But again this may require tightening some thresholds.

Best regards
Pete

cymantren
Posts: 10
Joined: 11 Jan 2014, 13:59
First name(s): Alfred
Last name(s): Güthler
Affiliation: privat
Country: Germany

Re: ZMAT and Symmetry

Post by cymantren » 23 Aug 2019, 15:17

Dear Mr Taylor and other DALTON-Users,

thank you for your valuable suggestions, but I was already aware of that fact, that a JT-Effect could appear. My problem in this case is to distinguish between an JT-effect and the reduction of the D3h-group to C2v by the the program. However, you wrote on the use of symmetry in the DALTON -manual and I have to read this more deeply to understand the use of Generators correctly.

To be more concrete, of course I accept the labeling of an 3-atomic Group: 2A_1+B_2, (C2v).
The Keyword "SUBSYM" is new to me. Does it prevent an "automatic" reduction of the D3h-group AND keeps the calculation to be C2v, if it is really an JT-Effect, otherwise it would be D3h ? Understanding this fact correctly, is essential to me, because I need to know if the reduction of symmetry is due to the JT-effect, or due to using a subgroup of D3h.
This effect appears only if degenerate orbitals are not equal occupied. I looked at the HOMO-LUMO gap and saw that there are NO degenarte orbitals. Of course C2v has no degenerate orbitals, however if the energy are almost equal, despite the fact, that the symmetry is different, is this then an hint for the possibility of an JT-Effect ?

It would be very kind to get answer to this.

Thank you !

Kind and best greetings

Alfred

Post Reply

Who is online

Users browsing this forum: No registered users and 5 guests