I think Turbomole can be added to the list of quantum chemistry programs that are interfaced to xtb. It's available in version 7.4.1.

document keywords introduced in grimme-lab/xtb#22

1. Please remove question marks:

• ".... with orca (??????????)."

In the tutorials for MTD, it is mentioned that one should exclude all constrained atoms from the RMSD list (see Conformers of transition-states, for instance). However, in doing so, won't I loose some degrees of freedom that I actually want to bias, i.e. all bonds to these atoms, except the ones involved in the transition state, or even the angle between the three atoms involved in the transition state and so on?

Describe the bug
This would seem to be a bug with CREST rather than xtb, but there's no repo for that, and it might also relate to the GFN2 energy levels.

I've been looking at the computed ensemble entropies from CREST and they seem somewhat high. If I take a simple set of branching alkanes (i.e., adding free rotor CH3 groups)

• methane = 0.0 J/mol•K (correct - no conformers, all entropy should be vib / rot / translational)
• ethane = 9.134 J/mol•K ?
• propane = 18.269 J/mol•K (exactly double)
• iso-butane = 27.403 J/mol•K (triple)
• neo-pentane = 36.538 J/mol•K (4x)

To Reproduce
Steps to reproduce the behaviour: Run any of the attached xyz files with default CREST options.

crest-issue.tar.gz

Expected behaviour

Since ethane, propane, butane have freely rotatable bonds, shouldn't the entropy scale with the gas constant (i.e., 8.314.. J/mol•K)?

I receive the error in the title and below any time I try and run a -hess job.

OpenBLAS Warning : Detect OpenMP Loop and this application may hang. Please rebuild the library with USE_OPENMP=1 option.

xtb was installed via conda (conda install xtb) and my current environment is:

name: xtb
channels:

• conda-forge
• defaults
  dependencies:
• _libgcc_mutex=0.1=conda_forge
• _openmp_mutex=4.5=0_gnu
• ca-certificates=2020.6.20=hecda079_0
• cairo=1.16.0=h3fc0475_1004
• certifi=2020.6.20=py38h32f6830_0
• fontconfig=2.13.1=h1056068_1002
• freetype=2.10.2=he06d7ca_0
• gettext=0.19.8.1=hc5be6a0_1002
• glib=2.65.0=h6f030ca_0
• icu=67.1=he1b5a44_0
• ld_impl_linux-64=2.34=h53a641e_5
• libffi=3.2.1=he1b5a44_1007
• libgcc-ng=9.2.0=h24d8f2e_2
• libgomp=9.2.0=h24d8f2e_2
• libiconv=1.15=h516909a_1006
• libpng=1.6.37=hed695b0_1
• libstdcxx-ng=9.2.0=hdf63c60_2
• libuuid=2.32.1=h14c3975_1000
• libxcb=1.13=h14c3975_1002
• libxml2=2.9.10=h72b56ed_1
• ncurses=6.1=hf484d3e_1002
• openbabel=3.1.1=py38hf4b5c11_0
• openssl=1.1.1g=h516909a_0
• pcre=8.44=he1b5a44_0
• pip=20.1.1=py_1
• pixman=0.38.0=h516909a_1003
• pthread-stubs=0.4=h14c3975_1001
• python=3.8.3=cpython_he5300dc_0
• python_abi=3.8=1_cp38
• readline=8.0=hf8c457e_0
• setuptools=47.3.1=py38h32f6830_0
• sqlite=3.30.1=hcee41ef_0
• tk=8.6.10=hed695b0_0
• wheel=0.34.2=py_1
• xorg-kbproto=1.0.7=h14c3975_1002
• xorg-libice=1.0.10=h516909a_0
• xorg-libsm=1.2.3=h84519dc_1000
• xorg-libx11=1.6.9=h516909a_0
• xorg-libxau=1.0.9=h14c3975_0
• xorg-libxdmcp=1.1.3=h516909a_0
• xorg-libxext=1.3.4=h516909a_0
• xorg-libxrender=0.9.10=h516909a_1002
• xorg-renderproto=0.11.1=h14c3975_1002
• xorg-xextproto=7.3.0=h14c3975_1002
• xorg-xproto=7.0.31=h14c3975_1007
• xz=5.2.5=h516909a_0
• zlib=1.2.11=h516909a_1006

Any suggestions on how to correct this problem? Thanks in advance.

XTB's MTD calculation is very powerful, especially Nanoreactor, because it can quickly respond to some large models. Examples of MTDs (such as Figure 6-11) in the literature (DOI: 10.1021 / acs.jctc.9b00143) can be placed in xtb_docs, or their calculation process files are provided in the software package.

I am trying to write to disk the spin density for a free radical, but when I activate the spinpol option xtb won't do so. Here is my command line:

xtb --input xtb.inp --molden radicale_propile-opt.xyz --spinpol --tblite --uhf 2

and my input file (xtb.inp):

$write
json=true
spin density=true
$cube
step=0.200

Github won't allow me to attach my xyz file, so here it is:

8

C -3.29241 3.36721 -0.04559
C -2.74829 1.97439 0.26550
C -1.22804 1.95824 0.26911
H -3.33840 1.14811 0.45005
H -0.70234 2.82495 0.07553
H -0.72180 1.08056 0.46514
H -4.36209 3.34171 -0.03990
H -2.94816 3.67740 -1.01005

Kind regards,

Silvio

Hi there,

I have one question about the --uhf in xtb. In most other programs, they use multiplicity to represent the electron spin information. But the definition here of uhf is a little different. it is number of unpaired electrons
When I tried to use xtb and xtbiff to generate interactions between CH3+ and indole. it seems that the E_ES_atom is really high in positive. Could you check if it is reasonable?
xtb ch3+.xyz --chrg 1 --uhf 0 --lmo
rm xtblmoinfo 1
xtb benzene.xyz --chrg 1 --uhf 0 --lmo
rm xtblmoinfo 2
xtbiff 1 2 -sp
By using this method, I got the information below (kcal/mol):
E Pauli : 3.132
E disp ATM: 0.001
E disp 2B: -1.430
E disp total : -1.428
E ES atom: 17.383
E ES LMO: 0.223
E ES total: 17.607
E induction: 0.022
E CT: -101.289
Eint total,gas: -81.957

Best,
Wentong

https://xtb-docs.readthedocs.io/en/latest/contents.html#xtb-in-other-quantum-chemistry-programs

cp2k has an xtb implementation since version 7.1
https://www.cp2k.org/version_history

Dear developers,
I find some new features have been released without updating the documents.

1. Implicit Solvation
   all available solvents have been extended to GFN-FF method since 6.3.0;
   implicit solvation model GBSA for GFN0-xTB seems to be supported since 6.2.

2. Detailed Input
   currently there are automatic check and remove of duplicated atoms from the fixing or constraining list as well as sorting of the list ?

3. Geometry Optimization
   improved defaults for optimizer (RF based for less than 500 atoms, L-BFGS based otherwise) since 6.3.0.

4. Periodic Boundary Conditions
   GFN1-xTB available under periodic boundary conditions since 6.3.0.

   Finally, I sincerely hope that some examples of pathfinder and nanoreactor using metadynamics coud be included in the documents.

Dear xTB developer:
Thank you very much for taking the time to read my error.
I want to use the pbc of xtb6.3.2.
According to the instructions of the help document, I created a "coord" file

$periodic 3
$cell angs
3.570 3.570 3.570 90 90 90
$coord angs
0.00000 0.00000 0.00000 C
0.89250 0.89250 0.89250 C
1.78500 1.78500 0.00000 C
2.67750 2.67750 0.89250 C
1.78500 0.00000 1.78500 C
2.67750 0.89250 2.67750 C
0.00000 1.78500 1.78500 C
0.89250 2.67750 2.67750 C
$end

and entered the following command:

xtb coord

Error information shows as follows:

[ERROR] Program stopped due to fatal error
-3- Single point calculation terminated
-2- xtb_calculator_singlepoint: Electronic structure method terminated
-1- scf: Multipoles not available with PBC
abnormal termination of xtb

I want to know if my input command is correct?
And if I want to do pbc structural optimization and molecular dynamics, can I directly add the following contents to the "coord" file?

$periodic 3
$cell angs
3.570 3.570 3.570 90 90 90
$coord angs
0.00000 0.00000 0.00000 C
0.89250 0.89250 0.89250 C
1.78500 1.78500 0.00000 C
2.67750 2.67750 0.89250 C
1.78500 0.00000 1.78500 C
2.67750 0.89250 2.67750 C
0.00000 1.78500 1.78500 C
0.89250 2.67750 2.67750 C
$opt
engine=inertial
$end

What should I do to solve this problem?

Looking forward to your reply, thanks~

My group has been looking to re-parametrize GFN2-xTB to calculate the vibrational structure of a specific class of small molecules. We would be happy to contribute a section to documents about how to do this since it appears to be missing. I am not sure if the omission is intentional but if it is desired we can add that section to the documentation if desired.

The global optimization program for chemical clusters, ABCluster (http://www.zhjun-sci.com/software-abcluster-download.php), has xTB interfaces implemented since version 1.5 (2017). I would appreciate it if you could have this mentioned in https://xtb-docs.readthedocs.io/en/latest/contents.html .

Could you please include the list of solvent names as they are accepted by XTB in the documentation.
I have tested them in XTB Version 6.1 beta, and saw that 'water' is not an accepted solvent name, only 'h2o' is. Also, 'n-Hexan' has to be written as 'nhexan' to be accepted. Of course no capital letters are also accepted. While most people can figure these things out through trial and error I see no reason why it shouldn't be documented.

The aISS program looks fantastic. I just tried the first example in the documentation and the lowest energy structure looks surprising to me (output attached). Several covalent bonds have broken and rearranged from the two input structures to give a covalent adduct plus trifluoromethane. I wanted to double check if the behavior is reproducible on your end? If so, is this a bug or a limitation of the model or perhaps I am missing something? Thanks for the awesome program. I am eager to try it out in other systems.

final_structures.txt

Since xtb4stda and stda launched on GH we should probably include a short guide on excited states in the documentation. Mainly to distinguish it from xtb and refer to the actual stda documentation.

@liljay42, @mdewergi would be nice if you could help out here, thanks.

https://xtb-docs.readthedocs.io/en/latest/optimization.html#example-2-cyclopentadienyl-anion

 xtb coord --opt tight --cycles 50 --charge -1

the example command line needs to be corrected to --chrg, instead of --charge

I actually would like to have also the additional optional option --charge because its more readable

I am trying to install xTB locally following the instructions described here: https://xtb-docs.readthedocs.io/en/latest/setup.html
I ran the following:

conda config --add channels conda-forge
conda install xtb

However, I cannot install it. I am getting the attached message. Does anyone know how to resolve this issue?
Thanks

Hello,
I need to run crest research with the constrain file and after a long try and error, I found that the indexing of the atoms starts at 1. Shall I suggest adding this information in the docs?

thanks so much

Hello,

1. add information that xtb and crest executables are open source and available from
   https://github.com/grimme-lab/xtb/releases

2. also add info how to unpack (tar xz) and setup.
   "tar xf xtb-191025.tar.xz"

3. Add config-xtb,bash for sourcing?
   There was an old shell script that could be sourced. Its not in the xtb-191025 release.

BTW, love the documentation, its very detailed.
Thank you!
Tobias

The xtb1 definition states that gbsa with 'benzene' is (only GFN1-xTB),

However the WIKI https://xtb-docs.readthedocs.io/en/latest/gbsa.html#parameterized-solvents states that benzene is also available for gfn2

solvents | GFN1 | GFN2
-- | -- | --
Acetone | x | x
Acetonitrile | x | x
Benzene | x | x

However executing

xtb aniline.xyz --opt extreme --gbsa benzene 
 Solvent             : benzene
#ERROR! solvent : benzene not parametrized for GFN2-xTB Hamiltonian
abnormal termination of xtb

using gfn1 it works as expected

 xtb aniline.xyz --opt extreme --gbsa benzene  --gfn1
normal termination of xtb

Solution: either parameterize benzene for gfn2 or change wiki

Dear XTB developers,

I'm trying to plot the IR spectrum of a molecule. I optimized the molecule and I calculated the frequencies (option -- hess).

If I understand correctly, to plot the spectrum I have to use the ir submodule with the --dftb hessian (obtained from the previous calculation) and --born charges (obtained from the previous calculation?). This is the command used:

xtb ir TPP_min.xyz --dftbplus hessian --born charges > out.log

And I obtain this error: " 'charges' file corrupted ... aborting"

Could you help me to solve the problem?

Attached file: file.zip

There was a bug regarding the calculation of f(0) with --vfukui. This bug will be fixed in the upcoming xtb version 6.2.

The BF3 example on the xtb_docs page already uses the xtb version with correct f(0).

This took me a long time to figure out and confused me very much, lots of pages mention "xcontrol(7)" but when searching the online web docs nothing shows up. So it would be nice to link to the man page or add the xcontrol parameter page to the website, because otherwise its not searchable via the little search box.

Example search for enan, should link to xcontrol, but yields nothing: https://xtb-docs.readthedocs.io/en/latest/search.html?q=enan&check_keywords=yes&area=default

The page https://xtb-docs.readthedocs.io/en/latest/xcontrol.html#
should link to here:
https://github.com/grimme-lab/xtb/blob/master/man/man7/xcontrol.7.txt

Hi,

I am trying to compare ENSO predicted NMR with experimental NMR taken from SDBSWeb (similar to what has been done in the tutorial). When I try to download the 1H NMR spectrum from SDBSWeb, the only downloadable data that I can find is the "peak data" which contains around 30 line of ppm, frequency, intensity data.

It seems nmrplot.py can not take these data as input(because a anmr.dat gives continues ppm vs intensity data). What exactly has been done in the tutorial to generate "exp.dat" so that we can compare experimental spectrum with enso predicted spectrum?

Hello. I have some problems with the usage of $path on xcontrol. I believe that it is a nudget-elastic band calculation (or some similar algorithm). Could I respectfully ask you to extend the corresponding information on this topic here, maybe with some example?

The Metadynamics Guide currently contains incomplete/wrong information.

• all duplicates with md.rst should be removed and references to md.rst instead
• the example given in mtd.rst will not perform a MTD calculation