Blog Archives

Geometry Optimizations for Excited States


Electronic excitations are calculated vertically according to the Frank—Condon principle, this means that the geometry does not change upon the excitation and we merely calculate the energy required to reach the next electronic state. But for some instances, say calculating not only the absorption spectra but also the emission, it is important to know what the geometry minimum of this final state looks like, or if it even exists at all (Figure 1). Optimizing the geometry of a given excited state requires the prior calculation of the vertical excitations whether via a multireference method, quantum Monte Carlo, or the Time Dependent Density Functional Theory, TD-DFT, which due to its lower computational cost is the most widespread method.

Most single-reference treatments, ab initio or density based, yield good agreement with experiments for lower states, but not so much for the higher excitations or process that involve the excitation of two electrons. Of course, an appropriate selection of the method ensures the accuracy of the obtained results, and the more states are considered, the better their description although it becomes more computationally demanding in turn.

Figure 1. The vertical excitation does not match the minimum on the excited state

In Gaussian 09 and 16, the argument to the ROOT keyword selects a given excited state to be optimized. In the following example, five excited states are calculated and the optimization is requested upon the second excited state. If no ROOT is specified, then the optimization would be carried out by default on the first excited state (Where L.O.T. stands for Level of Theory).

#p opt TD=(nstates=5,root=2) L.O.T.

Gaussian16 includes now the calculation of analytic second derivatives which allows for the calculation of vibrational frequencies for IR and Raman spectra, as well as transition state optimization and IRC calculations in excited states opening thus an entire avenue for the computation of photochemistry.

If you already computed the excited states and just want to optimize one of them from a previous calculation, you can read the previous results with the following input :

#p opt TD=(Read,Root=N) L.O.T. Density=Current Guess=Read Geom=AllCheck

Common problems. The following error message is commonly observed in excited state calculations whether in TD-DFT, CIS or other methods:

No map to state XX, you need to solve for more vectors in order to follow this state.

This message usually means you need to increase the number of excited states to be calculated for a proper description of the one you’re interested in. Increase the number N for nstates=N in the route section at higher computational cost. A rule of thumb is to request at least 2 more states than the state of interest. This message can also reflect the fact that during the optimization the energy ordering changes between states, and can also mean that the ground state wave function is unstable, i.e., the energy of the excited state becomes lower than that of the ground state, in this case a single determinant approach is unviable and CAS should be used if the size of the molecule allows it. Excited state optimizations are tricky this way, in some cases the optimization may cross from one PES to another making it hard to know if the resulting geometry corresponds to the state of interest or another. Gaussian recommends changing the step size of the optimization from the default 0.3 Bohr radius to 0.1, but obviously this will make the calculation take longer.

Opt=(MaxStep=10)

If the minimum on the excited state potential energy surface (PES) doesn’t exist, then the excited state is not bound; take for example the first excited state of the H2 molecule which doesn’t show a minimum, and therefore the optimized geometry would correspond to both H atoms moving away from each other indefinitely (Figure 2). Nevertheless, a failed optimization doesn’t necessarily means the minimum does not exist and further analysis is required, for instance, checking the gradient is converging to zero while the forces do not.

Figure 2. An unbound excited state with no minima ensures the dissociation of the system along the reaction coordinate

Basis Set Superposition Error (BSSE). A short intro


Molecular Orbitals (MOs) are linear combinations of Atomic Orbitals (AOs), which in turn are linear combinations of other functions called ‘basis functions’. A basis, or more accurately a basis set, is a collection of functions which obey a set of rules (such as being orthogonal to each other and possibly being normalized) with which all AOs are constructed, and although these are centered on each atomic nucleus, the canonical way in which they are combined yield delocalized MOs; in other words, an MO can occupy a large space spanning several atoms at once. We don’t mind this expansion across a molecule, but what about between two molecules? Calculating the interaction energy between two or more molecular fragments leads to an artificial extra–stabilization term that stems from the fact that electrons in molecule 1 can occupy AO’s (or the basis functions which form them) centered on atoms from molecule 2.

Fundamentally, the interaction energy of any A—B dimer, Eint, is calculated as the energy difference between the dimer and the separately calculated energies for each component (Equation 1).

Eint = EAB – EA – EB (1)

However the calculation of Eint by this method is highly sensitive to the choice of basis set due to the Basis Set Superposition Error (BSSE) described in the first paragraph. The BSSE is particularly troublesome when small basis sets are used, due to the poor description of dispersion interactions but treating this error by just choosing a larger basis set is seldom useful for systems of considerable sizes. The Counterpoise method is a nifty correction to equation 1, in which EA and EB are calculated with the basis set of A and B respectively, i.e., only in EAB a larger basis set (that of A and B simultaneously) is used. The Counterpoise method calculates each component with the AB basis set (Equation 2)

EintCP = EABAB – EAAB– EBAB (2)

where the superscript AB means the whole basis set is used. This is accomplished by using ‘ghost‘ atoms with no nuclei and no electrons but empty basis set functions centered on them.

In Gaussian, BSSE is calculated with the Counterpoise method developed by Boys and Simon. It requires the keyword Counterpoise=N where N is the number of fragments to be considered (for an A—B system, N=2). Each atom in the coordinates list must be specified to which fragment pertains; additionally, the charge and multiplicity for each fragment and the whole supermolecular ensemble must be specified. Follow the example of this hydrogen fluoride dimer.

%chk=HF2.chk
#P opt wB97XD/6-31G(d,p) Counterpoise=2

HF dimer

0,1 0,1 0,1
H(Fragment=1) 0.00 0.00 0.00
F(Fragment=1) 0.00 0.00 0.70
H(Fragment=2) 0.00 0.00 1.00
F(Fragment=2) 0.00 0.00 1.70

For closed shell fragments the first line is straightforward but one must pay attention that the first pair of numbers in the charge multiplicity line correspond to the whole ensemble, whereas the folowing pairs correspond to each fragment in consecutive order. Fragments do not need to be specified contiguously, i.e., you don’t need to define all atoms for fragment 1 and after those the atoms for fragment 2, etc. They could be mixed and the program still assigns them correctly. Just as an example I typed wB97XD but any other method, DFT or ab initio, may be used; only semiempirical methods do not admit a BSSE calculation because they don’t make use of a basis set in the first place!

The output provides the corrected energy (in atomic units) for the whole system, as well as the BSSE correction (which added to the previous term yields the un-corrected energy of the system). Gaussian16 also provides these values in kcal/mol as ‘Complexation energies’ first raw (uncorrected) and then the corrected energy.

BSSE is always present and cannot be entirely eliminated because of the use of finite basis sets but it can be correctly dealt with if the Counterpoise method is included.

Density Keyword in Excited State Calculations with Gaussian


I have written about extracting information from excited state calculations but an important consideration when analyzing the results is the proper use of the keyword density.

This keyword let’s Gaussian know which density is to be used in calculating some results. An important property to be calculated when dealing with excited states is the change in dipole moment between the ground state and any given state. The Transition Dipole Moment is an important quantity that allows us to predict whether any given electronic transition will be allowed or not. A change in the dipole moment (i.e. non-zero) of a molecule during an electronic transition helps us characterize said transition.

Say you perform a TD-DFT calculation without the density keyword, the default will provide results on the lowest excited state from all the requested states, which may or may not be the state of interest to the transition of interest; you may be interested in the dipole moment of all your excited states.

Three separate calculations would be required to calculate the change of dipole moment upon an electronic transition:

1) A regular DFT for the ground state as a reference
2) TD-DFT, to calculate the electronic transitions; request as many states as you need/want, analyze it and from there you can see which transition is the most important.
3) Request the density of the Nth state of interest to be recovered from the checkpoint file with the following route section:

# TD(Read,Root=N) LOT Density=Current Guess=Read Geom=AllCheck

replace N for the Nth state which caught your eye in step number 2) and LOT for the Level of Theory you’ve been using in the previous steps. That should give you the dipole moment for the structure of the Nth excited state and you can compare it with the one in the ground state calculated in 1). Again, if density=current is not used, only properties of N=1 will be printed.

Failure Reading NMR data in GaussView


There was this following message on a GIAO calculation when trying to open the file in GaussView5.0 (it opens successfully in ChemCraft)

CConnectionGLOG::Parse_GLOG()
Failure reading NMR data 
Line Number 2414

When you go to said line (line 2414) you find the following string:

Eigenvalues:-12345.6789 -12345.6789 -12345.6789

Which belong to the eigenvalues of the SCF NMR GIAO shielding tensor. The problem lies with the space missing between the colon sign ‘:’ and the ‘-‘ sign of the first eigenvalue. You can fix it either by hand with an editor but GV only warns you about the first instance so there may be others and you need to repeat the procedure. It is probably best to fix them all in one go with the following command from the terminal:

sed -i ‘s/Eigenvalues:-/Eigenvalues: -/g’

It is good to be back in Romania at the UBB writing these posts where this blog began. Thanks to my good friend Dr. Alexandru Lupan for pointing out this error.

Atom specifications unexpectedly found in input stream.


“Well, where else were they supposed to appear?”

I was sent this error along with the previous question for a failed optimization. Apparently there is no answer in the internet (I quickly checked) so here it is:

Gaussian is confused about finding atomic coordinates because there is also a geom=check instruction placed in the route section, i.e., it was told to retrieve the atomic coordinates from a checkpoint and then it was given those atomic coordinates within the input so it doesn’t know what you mean and exits.

Calculating NMR shifts – Short and Long Ways


Nuclear Magnetic Resonance is a most powerful tool for elucidating the structure of diamagnetic compounds, which makes it practically universal for the study of organic chemistry, therefore the calculation of 1H and 13C chemical shifts, as well as coupling constants, is extremely helpful in the assignment of measured signals on a spectrum to an actual functional group.

Several packages offer an additive (group contribution) empirical approach to the calculation of chemical shifts (ChemDraw, Isis, ChemSketch, etc.) but they are usually only partially accurate for the simplest molecules and no insight is provided for the more interesting effects of long distance interactions (vide infra) so quantum mechanical calculations are really the way to go.

With Gaussian the calculation is fairly simple just use the NMR keyword in the route section in order to calculate the NMR shielding tensors for relevant nuclei. Bear in mind that an optimized structure with a large basis set is required in order to get the best results, also the use of an implicit solvation model goes a long way. The output displays the value of the total isotropic magnetic shielding for each nucleus in ppm (image taken from the Gaussian website):

Magnetic shielding (ppm):
  1  C    Isotropic =    57.7345   Anisotropy =   194.4092
   XX=    48.4143   YX=      .0000   ZX=      .0000
   XY=      .0000   YY=   -62.5514   ZY=      .0000
   XZ=      .0000   YZ=      .0000   ZZ=   187.3406
  2  H    Isotropic =    23.9397   Anisotropy =     5.2745
   XX=    27.3287   YX=      .0000   ZX=      .0000
   XY=      .0000   YY=    24.0670   ZY=      .0000
   XZ=      .0000   YZ=      .0000   ZZ=    20.4233

Now, here is why this is the long way; in order for these values to be meaningful they need to be contrasted with a reference, which experimentally for 1H and 13C  is tetramethylsilane, TMS. This means you have to perform the same calculation for TMS at -preferably- the same level of theory used for the sample and substract the corresponding values for either H or C accordingly. Only then the chemical shifts will read as something we can all remember from basic analytical chemistry class.

GaussView 6.0 provides a shortcut; open the Results menu, select NMR and in the new window there is a dropdown menu for selecting the nucleus and a second menu for selecting a reference. In the case of hydrogen the available references are TMS calculated with the HF and B3LYP methods. The SCF – GIAO plot will show the assignments to each atom, the integration simulation and a reference curve if desired.

The chemical shifts obtained this far will be a good approximation and will allow you to assign any peaks in any given spectrum but still not be completely accurate though. The reasons behind the numerical deviations from calculated and experimental values are many, from the chosen method to solvent interactions or basis set limitations, scaling factors are needed; that’s when you can ask the Cheshire Cat which way to go

If you don’t know where you are going any road will get you there.

Lewis Carroll – Alice in Wonderland

Well, not really. The Chemical Shift Repository for computed NMR scaling factors, with Coupling Constants Added Too (aka CHESHIRE CCAT) provides with straight directions on how to correct your computed NMR chemical shifts according to the level of theory without the need to calculate the NMR shielding tensor for the reference compound (usually TMS as pointed out earlier). In a nutshell, the group of Prof. Dean Tantillo (UC Davis) has collected a large number of isotropic magnetic shielding values and plotted them against experimental chemical shifts. Just go to their scaling factors page and check all their linear regressions and use the values that more closely approach to your needs, there are also all kinds of scripts and spreadsheets to make your job even easier. Of course, if you make use of their website don’t forget to give the proper credit by including these references in your paper.

We’ve recently published an interesting study in which the 1H – 19F coupling constants were calculated via the long way (I was just recently made aware of CHESHIRE CCAT by Dr. Jacinto Sandoval who knows all kinds of web resources for computational chemistry calculations) as well as their conformational dependence for some substituted 2-aza-carbazoles (fig. 1).

1-s2.0-S0022286018310330-fx1_lrg

Journal of Molecular Structure Vol 1176, 15 January 2019, Pages 562-566

The paper is published in the Journal of Molecular Structure. In this study we used the GIAO NMR computations to assign the peaks on an otherwise cluttered spectrum in which the signals were overlapping due to conformational variations arising from the rotation of the C-C bond which re-orients the F atoms in the fluorophenyl grou from the H atom in the carbazole. After the calculations and the scans were made assigning the peaks became a straightforward task even without the use of scaling factors. We are now expanding these calculations to more complex systems and will contrast both methods in this space. Stay tuned.

Error for Gaussian16 .log files and GaussView5


There’s an error message when opening some Gaussian16 output files in GaussView5 for which the message displayed is the following:

ConnectionGLOG::Parse_Gauss_Coord(). 
Failure reading oriented atomic coordinates. Line Number

We have shared some solutions to the GaussView handling of *chk and *.fchk files in teh past but never for *.log files, and this time Dr. Davor Šakić from the University of Zagreb in Croatia has brought to my attention a fix for this error. If “Dipole orientation” with subsequent orientation is removed, the file becomes again readable by GaussView5.

Here you can download a script to fix the file without any hassle. The usage from the command line is simply:

˜$ chmod 777 Fg16TOgv5
˜$ ./Fg16TOgv5 name.log

The first line is to change and grant all permissions to the script (use at your discretion/own risk), which in turn will take the output file name.log and yield two more files: gv5_name.log and and name.arch; the latter archive allows for easy generation of SI files while the former is formatted for GaussView5.x.

Thanks to Dr. Šakić for his script and insight, we hope you find it useful and if indeed you do please credit him whenever its due, also, if you find this or other posts in the blog useful, please let us know by sharing, staring and commenting in all of them, your feedback is incredibly helpful in justifying to my bosses the time I spent curating this blog.

Thanks for reading.

fchk file errors (Gaussian) Missing or bad Data: RBond


We’ve covered some common errors when dealing with formatted checkpoint files (*.fchk) generated from Gaussian, specially when analyzed with the associated GaussView program. (see here and here for previous posts on the matter.)

Prof. Neal Zondlo from the University of Delaware kindly shared this solution with us when the following message shows up:

CConnectionGFCHK::Parse_GFCHK()
Missing or bad data: Rbond
Line Number 1234

The Rbond label has to do with the connectivity displayed by the visualizer and can be overridden by close examination of the input file. In the example provided by Prof. Zondlo he found the following line in the connectivity matrix of the input file:

2 9 0.0

which indicates a zero bond order between atoms 2 and 9, possibly due to their proximity. He changed the line to simply

2

So editing the connectivity of your atoms in the input can help preventing the Rbond message.

I hope this helps someone else.

Dealing with Spin Contamination


Most organic chemistry deals with closed shell calculations, but every once in a while you want to calculate carbenes, free radicals or radical transition states coming from a homolytic bond break, which means your structure is now open shell.

Closed shell systems are characterized by having doubly occupied molecular orbitals, that is to say the calculation is ‘restricted’: Two electrons with opposite spin occupy the same orbital. In open shell systems, unrestricted calculations have a complete set of orbitals for the electrons with alpha spin and another set for those with beta spin. Spin contamination arises from the fact that wavefunctions obtained from unrestricted calculations are no longer eigenfunctions of the total spin operator <S^2>. In other words, one obtains an artificial mixture of spin states; up until now we’re dealing only with single reference methods. With each step of the SCF procedure the value of <S^2> is calculated and compared to s(s+1) where s is half the number of unpaired electrons (0.75 for a radical and 2.0 for triplets, and so on); if a large deviation between these two numbers is found, the then calculation stops.

Gaussian includes an annihilation step during SCF to reduce the amount of spin contamination but it’s not 100% reliable. Spin contaminated wavefunctions aren’t reliable and lead to errors in geometries, energies and population analyses.

One solution to overcome spin contamination is using Restricted Open Shell calculations (ROHF, ROMP2, etc.) for which singly occupied orbitals is used for the unpaired electrons and doubly occupied ones for the rest. These calculations are far more expensive than the unrestricted ones and energies for the unpaired electrons (the interesting ones) are unreliable, specially spin polarization is lost since dynamical correlation is hardly accounted for. The IOP(5/14=2) in Gaussian uses the annihilated wavefunction for the population analysis if acceptable but since Mulliken’s method is not reliable either I don’t advice it anyway. 

The case of DFT is different since rho.alpha and rho.beta can be separated (similarly to the case of unrestricted ab initio calculations), but the fact that both densities are built of Kohn-Sham orbitals and not true canonical orbitals, compensates the contamination somehow. That is not to say that it never shows up in DFT calculations but it is usually less severe, of course for the case of hybrid functional the more HF exchange is included the more important spin contamination may become. 

So, in short, for spin contaminated wavefunctions you want to change from restricted to unrestricted and if that doesn’t work then move to Restricted Open Shell; if using DFT you can use the same scheme and also try changing from hybrid to pure orbitals at the cost of CPU time. There is a last option which is using spin projection methods but I’ll discuss that in a following post. 

Grimme’s Dispersion DFT-D3 in Gaussian #CompChem


I was just asked if it is possible to perform DFT-D3 calculations in Gaussian and my first answer was to use the following  keyword:

EmpiricalDispersion=GD3

which is available in G16 and G09 only in revision D, apparently.

There are also some overlays that can be used to invoke the use dispersion in various scenarios:

IOp(3/74=x) Exchange and Correlation Potentials

-77

-76

-60

-59

DSD-PBEP86 (double hybrid, DFT-D3).

PW6B95-D3.

B2PLYP-D3 (double hybrid, DFT-D3).

B97-D (DFT-D3).

IOp(3/76=x) Mixing of HF and DFT.

-33 PW6B95 and PW6B95-D3 coefficients.

IOp(3/124=x) Empirical dispersion term.

30

40

50

Force dispersion type 3 (Grimme DFT-D3).

Force dispersion type 4 (Grimme DFT-D3(BJ)).

Force dispersion type 5 (Grimme D3, PM7 version).

 

The D3 correction method of Grimme defines the van der Waals energy like:

$\displaystyle E_{\rm disp} = -\frac{1}{2} \sum_{i=1}^{N_{at}} \sum_{j=1}^{N_{at...
...{6ij}} {r_{ij,{L}}^6} +f_{d,8}(r_{ij,L})\,\frac{C_{8ij}} {r_{ij,L}^8} \right ),$

where coefficients $ C_{6ij}$ are adjusted depending on the geometry of atoms i and j. The damping D3 function for is:

$\displaystyle f_{d,n}(r_{ij}) = \frac{s_n}{1+6(r_{ij}/(s_{R,n}R_{0ij}))^{-\alpha_{n}}},$

where the values of s are adjustable parameters fit for the exchange-correlation functionals used in each calculation.

<span>%d</span> bloggers like this: