Quantics Exercises

Eryn Spinlove and Graham Worth

October 2024
Version: 1.1
 QUANTICS

A suite of programs for molecular QUANTum dynamICS simulations

Worth Group
University College London
Department of Chemistry

Quantics Exercises

Eryn Spinlove and Graham Worth

October 2024
 Eryn Spinlove and Graham Worth
Quantics Exercises
Handout, October 2024
Contributors: The Worth Group

University College London

Department of Chemistry
Faculty of Mathematics and Physical Sciences
20 Gordon Street
London
WC1H 0AJ

iv
Abstract

Quantics is a general purpose program package to simulate the evolution of

a molecular system by solving the time-dependent Schrödinger equation.

The code is based around the multi-configurational time-dependent Hartree

(MCTDH) algorithm in various variants, including the powerful multilayer-
MCTDH algorithm that has been used to propagate a wavefunction for up to
1000 degrees of freedom. MCTDH uses a contraction of basis set representa-
tions of the Hamiltonian and wavefunction, and Quantics includes a range
of variable representation (DVR) grid basis sets and collocation methods.

Input is via ascii text files and for molecules with analytical potential func-
tions no programming is required. A library of potential functions is included
to treat more complicated cases, and further functions can be added as
required by the user.

The code also includes the variational multi-configurational Gaussian

(vMCG) method that is based on a Gaussian wavepacket expansion of the
wavefunction. vMCG can be run in a “direct” manner (DD-vMCG), calcu-
lating the potential energy surfaces on-the-fly using a number of quantum
chemistry programs. In addition to wavepacket propagation, Quantics can
solve the time-independent Schrödinger equation for small systems and can
solve the Liouville–von-Neumann equation to propagate density matrices.
The Package includes auxiliary programs to help set up calculations and
analyse the output.

Quantics is a community code of the UK Collaborative Computational Project

for Quantum Dynamics (CCPQ) and the European E-CAM project, an e-
infrastructure for software development run by the Centre Européen de
Calcul Atomique et Moléculaire (CECAM). Through this it has become a
framework for general dynamics codes, for example enabling an external
surface hopping code to use the Quantics input and operator interfaces.

v
 More details of the code and its capabilities are in Worth Comp. Phys.
Comm.(2020) 248, 107040.

vi
Contents

1 Introduction 1
1.1 Starting Out . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 What’s In This Handout . . . . . . . . . . . . . . . . . . . . . . . 2

2 Exercise 1: One-Dimensional Oscillator 3

2.1 First Calculations: The Harmonic Oscillator . . . . . . . . . . . 3
2.1.1 Operator File . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.2 Input File . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1.3 Running the Calculation . . . . . . . . . . . . . . . . . . 4
2.1.4 Analysing the Results . . . . . . . . . . . . . . . . . . . . 5
2.1.5 Changing the system . . . . . . . . . . . . . . . . . . . . 7
2.1.6 Anharmonicity: The Morse Oscillator . . . . . . . . . . . 8
2.2 Tully Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.1 Simple Avoided Crossing (Tully 1) . . . . . . . . . . . . . 9
2.2.2 Dual Avoided Crossing (Tully 2) . . . . . . . . . . . . . . 12
2.2.3 Extended Coupling with Reflection (Tully 3) . . . . . . . 13

3 Exercise 2: Non-adiabatic Dynamics 15

3.1 The Butatriene Cation . . . . . . . . . . . . . . . . . . . . . . . . 15
3.2 Simulating the Dynamics . . . . . . . . . . . . . . . . . . . . . . 16
3.2.1 Absorption Spectrum . . . . . . . . . . . . . . . . . . . . 18
3.2.2 Expectation Values . . . . . . . . . . . . . . . . . . . . . . 19
3.3 Stick Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

4 Exercise 3: vMCG 21
4.1 Using a GWP basis . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.2 Surface Hopping . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

5 Exercise 4: DD-vMCG 25
5.1 Butatriene Cation using Molpro . . . . . . . . . . . . . . . . . . 25
5.2 Water Excitation using Molpro . . . . . . . . . . . . . . . . . . . 28

6 Exercise 5: NOCl Photodissociation 31

6.1 The Absorption Spectrum . . . . . . . . . . . . . . . . . . . . . . 31
6.2 Computing excitation and reaction probabilities . . . . . . . . 34

vii
 Bibliography 39

Glossary 40

viii
Introduction 1
„ You can’t do better design with a computer,
but you can speed up your work
enormously.

— Wim Crouwel
(Graphic designer and typographer)

1.1 Starting Out

In the following are a set of example calculations to introduce the basic

ideas of quantum dynamics calculations and how they are performed using
the Quantics program.

The information given is reduced as the examples progress and looking back
to earlier examples should give the answer as to how something should
be done. Further information is to be found in the on-line documentation
included with the package, where detailed information on keywords available
etc. are present.

The Quantics website can be found here, or at

https://www.chem.ucl.ac.uk/quantics/index.html

There is a section titled Quantics Training Workshop, Nov. 2024 with

links to this manual and supporting files. You will need a copy of quantics
installed on your computer. If $QUANTICS_DIR is the root directory of
the Quantics installation, the documentation is to be found at $QUAN-
TICS_DIR/doc/index.html

1
 1.2 What’s In This Handout

This handout contains the exercises which should help you to successfully
run different types of calculations with the Quantics software package.

• Chapter 2 Setting up and running 1D harmonic oscillator system, in-

troduction to the analysis/plotting programs and running calculations
on the Tully model systems

• Chapter 3 Setting up and running non-adiabatic dynamics on the

Butatriene cation and generating spectra from the results

• Chapter 4 Setting up and running non-adiabatic dynamics using the

vMCG (Gaussian wavepacket) method

• Chapter 5 Setting up and running direct quantum dynamics simula-

tions using the DD-vMCG method

• Bibliography Key papers for the methodology/implementation of the

methods in each section.

• Glossary Key commands and terms with short descriptions

2 Chapter 1 Introduction
Exercise 1: One-Dimensional 2
Oscillator

„ A picture is worth a thousand words. An

interface is worth a thousand pictures.

— Ben Shneiderman
(Professor for Computer Science)

In the first set of exercises we will look at some simple one-dimensional

systems to learn the thinking process and basic input required for a Quantics
calculation.

2.1 First Calculations: The Harmonic Oscillator

The Hamiltonian for a 1D harmonic oscillator is

1 ∂2
H=− + 1 mω 2 q 2 (2.1)
2m ∂q 2 2

where m is the mass and ω the frequency. Note that we are using atomic
units in which h̄ = 1 so it can be omitted.

2.1.1 Operator File

To specify a Hamiltonian, you need an operator file. The operator file “ho.op”
codes the Harmonic oscillator Hamiltonian above. You will see there are 3
SECTIONS.

The OP_DEFINE-SECTION contains a title that will be printed in output.

The HAMILTONIAN-SECTION defines the operator. The line beginning

"modes" defines the fact there is a single coordinate, labelled "x". The 2 lines
between the dashes (these are not required for the operator, and are included
to help readability) define 2 terms that can be related to the Hamiltonian
2.2.

3
 The PARAMETER-SECTION defines the mass and frequency. Values of a
proton mass with a frequency close to 3500 cm−1 are given. This is an
approximate C–H stretch.

2.1.2 Input File

To run a calculation you need an input file. Look at ho.inp. Again there are
various SECTIONs.

The OPERATOR-SECTION defines the file containing the operator - here

"ho.op".

The PRIMITIVE-BASIS-SECTION defines the “primitive basis set” - a grid

of points on which the wavepacket will be propagated. Here 121 points
between x = −2 au and x = 2 au are defined. ( 1 au ∼ 0.5Å ).

The SPF-BASIS-SECTION defines the number of “single-particle” basis

functions that will be used to describe the evolving wavepacket. For a 1D
problem only 1 function is required.

The INIT_WF-SECTION defines the initial wavefunction. A Harmonic os-

cillator ground-state eigenfunction is specified here for mode x. The 4
parameters specify the centre of the wavepacket, its momentum, and the
frequency and mass of the harmonic oscillator. Notice that the oscillator
has been displaced to x = 2.0 au.

Finally, the RUN-SECTION defines what will be calculated. Here we are

running a propagation that will last for 20 fs with output written ever 0.5
fs. Output will be written to a directory “ho”.

2.1.3 Running the Calculation

Step 1. Set-up the directory and run the calculation

• mkdir ho
• quantics ho

It should finish in a few seconds.

Step 2. Move into the directory and list the contents

4 Chapter 2 Exercise 1: One-Dimensional Oscillator

 • cd ho
• ls

There will be a number of files including psi

2.1.4 Analysing the Results

Step 3. Plot the potential surface and wavepacket motion using the showsys
program. This is driven by a simple menu. Type the number to activate the
desired option and follow the instructions. You might have to type enter a
number of times to be able to enter an option. To start type

• showsys -gnupl4

The option -gnupl4 is needed to use the verion of gnuplot installed on

the cluster. A menu will appear. Look at item 10, the plot task. Type

• 10

A menu will appear with the possible plot tasks. Select

• 2 plot pes

Now type 1. This activates “plot to screen” and a plot of the potential
surface will be shown. This should be a quadratic curve. Try changing
the energy units using option 90 to cm− 1 by using the keyword cm-1.
Try changing the range over which the PES is plotted using options 30 (to
change x-axis) and 70 (to change y-axis).

Step 4. Now plot the wavepacket motion. From option 10 choose

• 7 plot reduced density

The density is Ψ2 , i.e. the probability density for the particle. Reduced
density means that the density has been integrated over any other degrees
of freedom to allow the 1D plot to be made. Type option 280 (this is not
visible in the menu), select

• 2 = movie

2.1 First Calculations: The Harmonic Oscillator 5

 and then type 1, it will provide a movie of the motion of the wavepacket.

The wavepacket will move and the time shown at the top of the plot will be
incremented. The plots should be 0.5 fs apart - the value of tpsi in the
RUN-SECTION. If this was given a different value the snapshots would be
at this new value.

It is also possible to view the motion of the wavepacket as stepwise snapshots

by skipping using option 280 and directly typing 1.

Step 5. Plot the trajectory of the centre of the wavepacket. The easiest way
to do this is to use another program that will extract the expectation value
of the coordinates stored in the file check. Type

• rdcheck -g qdq 1 1

This will produce a file qdq_1_1.pl for the first mode of the first (and only)
state. The option -g meant that this file is ready to be plotted using the
gnuplot program. Type

• gnuplot

A prompt gnuplot> appears. To plot column 1 against 2 of the data using

a line plot, type

• plot "qdq_1_1.pl" u 1:2 w l

This is the expectation value of the coordinate, which can be interpreted

as the centre of the wavepacket. It should oscillate from 2.0 to -2.0, and
should agree with what you saw as the wavepacket motion.

The time-period of a harmonic oscillator is given by

2π
T =
ω

If T is in fs and ω in cm−1 , then using the appropriate conversions factors

33356
T =
ω

Consider This: For the system above what is the time period? Is that what
is seen?

6 Chapter 2 Exercise 1: One-Dimensional Oscillator

Column 3 contains the width of the wavepacket, defined as the standard
√
deviation < x2 > − < x >2 . This is plotted using

• plot "qdq_1_1.pl" u 1:3 w l

It also oscillates, but looking at the values is nearly constant.

2.1.5 Changing the system

To run a similar, but altered, calculation move back up to the starting

directory and copy the operator and input files.

• cd ..
• cp ho.op ho1.op
• cp ho.inp ho1.inp

Change the frequency of the oscillator in ho1.op to give a time period of 35

fs (change the parameter in the operator file). Run a propagation of 70 fs
(2 time periods) by changing tfinal in the new input file. Also change the
value of the name variable so that all the results are written to a directory
ho1. Now make the new directory and run the new calculation

• mkdir ho1
• quantics ho1

Change directory into ho1 and use showsys -gnupl4 to plot the potential
surface and wavepacket motion as before.

Consider This: Is your result what you would expect?

To the see the effect of different starting conditions, first try altering the
starting points. To do this, change the centre of the initial wavepacket
in ho1.inp (the first parameter in the line defining the initial wavepacket
in the INIT_WF-SECTION). Also try different initial widths (change the
frequency - the third parameter in the line defining the initial wavepacket
in the INIT_WF-SECTION).

Points to note:

• The time period does not depend on the initial displacement.

2.1 First Calculations: The Harmonic Oscillator 7

 • If the frequency defining the initial wavepacket width matches the
frequency of the potential function the wavepacket remains the same
width at all times.

2.1.6 Anharmonicity: The Morse Oscillator

Finally we shall change the harmonic oscillator to a Morse oscillator. To

replace the harmonic potential with a Morse function. involves replacing
the 12 ωq 2 operator with a label such as V and defining the Morse operator in
a LABELS-SECTION. We will base the operator on the harmonic operator
file ho.op

• cp ho.op mo.op

Now edit mo.op.

• In the PARAMETER-SECTION add parameters for the system mass (we

will use mass-scaled coordinates) and the well depth.

depth = 8.0, ev
mass_x = 1.0, amu

• In the HAMILTONIAN-SECTION replace the KE and potential operators

with

1.0 | KE
1.0 | V

• The label V now needs to be defined by adding a LABELS-SECTION

LABELS-SECTION
V = morse[depth, omega, 0.0, 0.0, mass_x]
end-labels-section

• mkdir mo

Using the input file mo.inp run and look at the dynamics. Note that it looks
much like the harmonic systems. Now move the initial wavepacket out, first
to be centred at -0.2 and then -0.3. Notice the effects of the anharmonicity.

8 Chapter 2 Exercise 1: One-Dimensional Oscillator

2.2 Tully Models

In the next set of exercises we will look at some simple scattering problems
with the Tully Models. These were set up by John C. Tully [3] to provide
models to test the performance of surface hopping calculations in including
non-adiabatic effects in scattering calculations.

The Hamiltonian for the Tully Models is written in a diabatic basis:

!
1 ∂2 V11 V12
H=− + (2.2)
2m ∂q 2 V21 V22

with m is the mass and is set to 2000. Note that we are using atomic units
in which h̄ = 1 so it can be omitted. As a comparison, a proton in atomic
units has a mass of 1822.

2.2.1 Simple Avoided Crossing (Tully 1)

The operator file is tully_scm.op and tully_scm.inp the input file. The
potentials are:

V11 = A (1 − exp(−Bx)) , x>0

V11 = −A (1 − exp(Bx)) , x<0
V22 = −V11
V12 = V21 = C exp(−Dx2 )

with empirical paramaters A, B, C, D.

Step 1. Set-up the directory and run the calculation

• mkdir tully_scm
• quantics tully_scm

It should finish in a few seconds.

Move into the directory and analyse the results.

Step 2. Plot the potential surface and wavepacket motion utilising the
showsys program using the same options as for the harmonic oscillator
(Section 2.1.4). There are now 2 states and the state of interest can be
selected by

2.2 Tully Models 9

 • 60 = change state selection ( 1)

Type 60. In response to the question“How many states will be plotted?”

select 1. Then in response to “Enter selection” enter 2. The wavepacket or
potential can now be plotted for the upper state.

Type 60 again. In response to the question“How many states will be plot-

ted?” now select 2. Then in response to “Enter selection” enter “1 2”. The
wavepacket or potential can now be plotted for both states.

You should see that after 60fs the crossing region is reached and population
transfers to the state 2 state where it passes on the end of the grid, while on
state 1 it relects back in the incoming channel. You will also see that at later
times there is refection from the end of the grid that causes interference
with the outgoing wave.

Step 3. To remove this interference, edit the file tully_scm.inp to comment

in

#alter-labels
#CAPl_x = CAP[-12.0 0.0001 3 -1]
#CAPr_x = CAP[12.0 0.0001 3]
#end-alter-labels

i.e. remove the “#” at the start of each line. This will add CAPs at the ends
of the grid. Rerun the calculation

• quantics -w tully_scm

the “-w” option overwrites the previous results. If you now plot the wavepacket
it will be absorbed before reaching the end and the interference will not be
present.

Step 4. Analyse the branching ratio, which is percentage of the wavepacket

that is transmitted through the avoided crossing (exits to RHS of the grid)
divided by the percentage of the wavepacket that is reflected (exits to LHS of
grid). To do this we integrate the flux going into the CAP over the propagation
using the program flux. This program has many possible uses - type flux
-h to see all the options.

First go into the directory with results

10 Chapter 2 Exercise 1: One-Dimensional Oscillator

 • cd tully_scm

Now run flux using the option -wtt to calculate the expectation of the
flux going into the CAP at the left hand end of the x coordinate (reflection).
Save the output files wtt and iwtt that contain the flux and integrated
flux, respectively.

• flux -wtt x left

• mv wtt wtt.left
• mv iwtt iwtt.left

Do the same for the flux going into the CAP at the right hand end of the x
coordinate (transmission).

• flux -wtt x right

• mv wtt wtt.right
• mv iwtt iwtt.right

Now plot the results using gnuplot. First look at the fluxes.

• gnuplot
• plot "wtt.left" u 1:2 w l title "reflected"
• replot "wtt.right" u 1:2 w l title "transmitted"

These should go to zero at long times to indicate the scattering has finished.
Now look at the integrated flux.

• plot "iwtt.left" u 1:2 w l title "reflected"

• replot "iwtt.right" u 1:2 w l title "transmitted"

You will see that the integrated fluxes reach a plateau. The final values give
the branching ratio of transmitted/reflected ≈ 0.55/0.45.

Step 5. Repeat the calculation with different initial momenta. For each
calculation copy the input file to a new name, e.g. cp tully_scm.inp
tully_scm1.inp . Edit this new file

• change the name directory so the data is put in a new location, e.g.
name = tully_scm1.

2.2 Tully Models 11

 • change the momenta in the INIT_WF-SECTION by editing the number
4.5 in the line:
x Gauss -8.0 4.5 0.5

• Make the new directory mkdir tully_scm1 and run the new calcula-
tion.

If you look in the documentation, the 3 numbers in the initial wavepacket

definition define the position, momentum and width, respectively.

Repeat the analysis and see if more or less wavepacket is transmitted.

2.2.2 Dual Avoided Crossing (Tully 2)

The potentials for the Dual Avoided Crossing are defined as:

V11 = 0
V22 = −A exp(−Bx2 ) + E0
V12 = V21 = C exp(−Dx2 )

with empirical paramaters A, B, C, D, E0 . Get the operator for this model,

tully_dac.op and set up the input file for a calculation

• cp tully_scm.inp tully_dac.inp

Edit the input file to use the different operator.

• change the name directory so the data is put in a new location, e.g.
name = tully_dac

• change the name of the operator file e.g. opname = tully_dac

• Make the new directory mkdir tully_dac and run the new calcula-
tion.

Analysis can be carried out as for the Tully Model 1.

12 Chapter 2 Exercise 1: One-Dimensional Oscillator

2.2.3 Extended Coupling with Reflection (Tully 3)

Repeat the calculations for the Tully model 3, the Extended Coupling with
Reflection by copying and editing the input file tully_scm.inp. The poten-
tials are:

V11 = A
V22 = −A
V12 = V21 = B exp(Cx) , x<0
V12 = V21 = B (2 − exp(−Cx)) , x>0

with empirical paramaters A, B, C. This operator is coded in tully_ecwr.op.

2.2 Tully Models 13

Exercise 2: Non-adiabatic 3
Dynamics

„ Don’t start anything new on a Friday

afternoon.

— Graham Worth
(Professor of Computational Chemistry)

We shall now try a multi-dimensional, multi-state calculation that uses the

MCTDH algorithm. We shall also check the quality of the calculation in
addition to plotting the spectrum and analysing the dynamics.

3.1 The Butatriene Cation

The butatriene (C4 H+

4 ) is a classic example of a molecule with a conical
intersection between the two lowest states. The Hamiltonian to first-order
can be written using a vibronic coupling model as

(1)
! ! !
X ωi ∂2 2
X κi Qi 0 0 λQ5
H= + Qi + +
i=1,18
2 ∂Q2i i=8,12,14,15 0 (2)
κ Qi λQ5 0
(3.1)
Note that only 4 modes provide on-diagonal coupling and 1 mode off-diagonal
coupling. This Hamiltonian has been coded up in the file C4H4_linear.op
with the input file butatriene.inp.

In order to view the PESs, first run a test calculation using the -t option,
in which the calculation stops after calculating the initial energy.

• mkdir butatriene
• quantics -t butatriene
• cd butatriene
• showsys -gnupl4

15
 Option 20 is used to select the modes for the plot. Check that it says

• 20 = change coordinate section (v10a=x v6a=y)

If not, type 20 and use the option to select these modes. This option can be
used to select any 1D or 2D cut through the surfaces. Use menu item 60
to plot both electronic surfaces. Now use option 10 to select what is to be
plotted. We want to plot a pes. Look at the differences between the diabatic
and adiabatic pictures. The conical intersection is clear in the adiabatic
picture where the surfaces meet. The best plot is to set the surfaces (option
160) and the z-axis to be between 0 and 2.0 (option 70).

3.2 Simulating the Dynamics

Move back up to the starting directory and run the dynamics.

• quantics -w butatriene

This simulates a vertical excitation from the ground-state to the upper state.
Move back into the name directory butatriene to analyse the calculation.

First, we need to check that the calculation is of a good enough quality,

i.e. check for convergence of the basis sets. Start up the analysis program
rdcheck

• rdcheck natpop

The natural orbitals are an ordered representation of the SPFs and for
a good calculation the population of the highest natural orbital must be
small, i.e. the basis function plays a small role in the wavefunction. This
calculation has 2 states and 2 modes. Enter 0 for the mode and the state
to be analysed, and the maximum population of the highest natural orbital
is printed. The highest orbitals have negligible populations, so the basis is
good.

Next check the “primitive grid” - we are using a DVR basis here so we will
actually be checking whether we have enough primitive basis functions to
describe the SPFs, but we loosely refer to this basis as the grid. Run the
program rdgpop . In answer to the question ““no. ==?” enter 1. This will
return the population of the grid points at either end of the grid. Entering

16 Chapter 3 Exercise 2: Non-adiabatic Dynamics

a larger number will provide a population summed over the selected no. of
points from the end. Select dof 0. The maximum population for the modes
are then printed to screen. The population along the first mode is quite
large indicating that the primitive basis set might be too small.

Now plot the diabatic state populations using the program statepop. In
the directory with the results files, type

• statepop -inter -gnupl4

Select option 1 - “Plot to screen”. The resulting plot shows how the system
begins in one of the components of the degenerate state and oscillates
rapidly between the two.

Next, plot the adiabatic populations. To do this we need to generate the

populations and use gnuplot to plot them. Type

• adpop

• gnuplot

In gnuplot, type

• plot "adp" u 1:2 w l

• replot "adp" u 1:3 w l

The resulting plot shows that the system starts with some initial population
in the lower adiabatic states. As time goes on, the population moves steadily
to the lower state.

Finally, plot the evolving wavepacket using the analysis program showsys.
To select the adiabatic or diabatic wavefunction use option 10. To select the
wavepacket on either state use option 60. The pictures give a similar result.
The wavepacket in the upper state remains compact and oscillates in the
upper cone along v6a. The wavepacket appears in the lower state at around
30 fs when the wavepacket has reached the intersection. It then oscillates
around the ground-state surface, spreading in both directions showing that
the torsion, v10a, has been excited by the internal conversion through the
intersection.

3.2 Simulating the Dynamics 17

 3.2.1 Absorption Spectrum

An absorption spectrum is calculated from the Fourier Transform of an

autocorrelation function, i.e.
Z ∞
I(ω) ∝ dt C(t)eiωt
−∞

where
C(t) = ⟨Ψ(0)|Ψ(t)⟩

Assuming a vertical excitation, similar to that in Fig. 3.1, we can simply

propagate the ground-state wavefunction placed “vertically” onto the ex-
cited state. The keyword init_state = 2 in the INIT_WF-SECTION of the
butatriene simulation is the command that places the wavepacket on the
upper state.

S2

- 

 
- S1
Energy

JJ hν

S0

Fig. 3.1: Schematic diagram of vertical excitation from S0 into S2

In order to plot the autocorrelation function we need the program rdauto.

Type

• rdauto -inter

A menu will appear (-inter is an argument that provides this). Plot the
absolute value of the autocorrelation function. This is the default plot task
and so this can be done by simply typing 1. This is a periodic function
with the time period of the harmonic motion that starts with a value of 1
(complete overlap) and falls to 0.8 as the wavepacket moves away.

To plot the spectrum, type

18 Chapter 3 Exercise 2: Non-adiabatic Dynamics

 • autospec -inter

You will be asked to enter E-min, E-max, unit =? (e.g: 2., 5., ev),
i.e. the energy bounds for the spectrum and the units. Enter 8.5 10.5 ev.
In response to tau =? (damping time in fs) respond with 30 and
for iexp =? (damping order) enter 1. Note that a larger damping time
leads to narrower peaks in the spectrum) and the damping order is the
exponent of the cutoff function that ensures the correlation function goes
to zero at time T.

3.2.2 Expectation Values

It is possible to obtain the expectation value of an operator of interest during

a simulation by coding the operator and using the expect keyword in the
RUN-SECTION. Here we look at the situation of how to obtain the expectation
value after the propagation has been run.

We want to obtain a measure of the energy flow out of mode Q14 in state
S2 to both S1 and to Q5 . For this we will calculate the expectation of the
operator for the 1D oscillator
!
ω14 ∂2 (2)
O= + Q214 + κi Q14 |2⟩⟨2| (3.2)
2 ∂Q214

To code this, add to the operator file C4H4_linear.op the following lines

HAMILTONIAN-SECTION_HOenergy
modes | el | Q_1 | Q_2 | Q_3 | Q_4
modes | Q_5 | Q_6 | Q_7 | Q_8 | Q_9
modes | Q_10 | Q_11 | Q_12 | Q_13 | Q_14
modes | Q_15 | Q_16 | Q_17 | Q_18

omega_14 |1 S2&2 |2 KE
0.5*omega_14 |1 S2&2 |2 q^2
kappa2_14 |1 S2&2 |15 q
end-hamiltonian-section

This will define an operator with the label HOenergy to distinguish it from
the system Hamiltonian that has no label.

3.2 Simulating the Dynamics 19

 We now need to add this operator file to the oper file so that it can be used
by Quantics. To do this, edit the input file:

• In the RUN-SECTION remove the keyword propagation and replace

it with the keyword genoper program.

now run

• quantics -w butatriene

Change into the output directory and run the expect program

• expect

A list of possible operator will be given. Select HOenergy. This will provide
a file.

3.3 Stick Spectrum

For small systems it is possible to solve the time-independent Schrödinger

equation and diagonalise the Hamiltonian to produce a stick spectrum of
the eigenvalues. Copy the file butatriene.inp to but_stick.inp. Replace
the RUN-SECTION with

RUN-SECTION
name = but_stick diagonalise = 50
end-run-section

and run the Quantics calculation.

Move into the results directory. There will be a file eigval. Plot the stick
spectrum using the analysis program rdeigval -inter. Compare the
resulting spectrum to that from the time-dependent calculation , the same
peaks should be present in both. The time-dependent spectrum could be
improved by a longer propagation time and longer damping factor in the
spectrum calculation to make the lines sharper.

20 Chapter 3 Exercise 2: Non-adiabatic Dynamics

Exercise 3: vMCG 4
„ Users do not care about what is inside the
box, as long as the box does what they
need done.

— Jef Raskin
about Human Computer Interfaces

The vMCG method replaces the single-particle functions of MCTDH with

Gaussian wavepackets (GWP). In the following exercises we will use this basis
type on anaytical potential surfaces to see the extra analysis available.

4.1 Using a GWP basis

Take the butatriene input file (butatriene.inp) and create a new input file,
butatriene_gwp.inp changing the name directory to butatriene_gwp for the
new output. Now change the primitive basis to a GWP basis. I.e in the
PRIMITIVE-BASIS-SECTION replace on all the lines

HO 10 0.0 1.0 1.0

with

GWP

Also change the 1-dimensional SPFs to be multi-dimensional GWPs by

replacing

SPF-BASIS-SECTION
multi-set
Q_5 = 8,8
Q_14 = 8,8
end-spf-basis-section

21
 to

SPF-BASIS-SECTION
single-set
Q_14, Q_5 = 8
end-spf-basis-section

The keyword single-set means that we do not use a separate set of func-
tions of functions for each state.

Run this calculation.

Plot the state populations using

• statepop -inter -i butatriene_gwp

and compare it to the state populations obtained from the MCTDH calcula-
tion

• statepop -inter -i butatriene

The GWP result has significant differences due to the small basis set size.
Rerun the calculation with more GWPs by editing the line Q_14, Q_5 =
8, increasing the basis set until convergence is reached. I.e. the state
populations match that of the MCTDH calculation.

To plot the trajectories at the centres of the wavepackets, use the analysis
program gwptraj

• gwptraj -i butatriene_gwp -gnupl4

4.2 Surface Hopping

Quantics is coupled to a surface hopping code [1] that uses Tully fewest
switches. To demonstrate ow it works we will use a simple 2-dimensional
model of a conical intersection from Ferretti [2].

Take the input file ferretti.inp and the associated operator ferretti.op.
This is a for a numerically exact simulation starting on the upper surface.

22 Chapter 4 Exercise 3: vMCG

Run this calculation. For analysis, plot the surfaces using showsys and then
the diabatic state populations using statepop. The population transfer is
at around 40 fs and almost complete.

Calculate and plot the adiabatic populations:

• cd ferretti
• adpop

and then use gnuplot

• plot "adp" u 1:3 w l

• replot "adp" u 1:4 w l

You will see that the transfer is minimal. If you look at the wavepacket (use
showsys) you will see that it all stays on the upper adiabatic surface and
the crossing is purely diabatic. Note that we are plotting the populations
for states 2 and 3 as the model includes the ground state (state 1).

Run the calculation again, but placing the initial wavepacket further away
from the intersection. To do this edit the INIT_WF-SECTION, replacing

• x HO 2.0 0 0.001 20000

with

• x HO 4.0 0 0.001 20000

The conical intersection is at x = 3.5 so this moves the wavepacket away.

As a result, the wavepacket has a higher momentum as it reaches the
crossing. Plot the diabatic and adiabatic populations. You will see that now
the diabatic transfer is less, but the adiabatic transfer is significant and
permanent.

Now take the input file ferretti_tsh.inp. This is for a surface hopping
calculation with 10 trajectories. It is a complicated input as we need to
transform the coordinates to Cartesian coordinates (the tshtrans-section
and provide the parameters for the surafce hopping code. These are in the
SH_ZAGREB-SECTION (the code is known as the “Zagreb Surface Hopping
program”). Here parameters such as the time step and random number

4.2 Surface Hopping 23

 generator seed are defined. Other than that, the input is similar to that for
an MCTDH calculation, with the same initial wavepacket definition.

The code samples this initial wavepacket using a Wigner distribution and
runs each trajectory in turn, storing them in a directory ferretti_tsh/zagreb_trj.
When they are all completed the code constructs the files needed for analysis.
To plot the state populations, use

• statepop -inter -gnupl4

as before. This time, adiabatic populations are plotted. They compare quite
well to the results from the MCTDH simulation starting at x = 0. For better
converegence, add some more trajectories by editing the keyword swarm =
10 in the RUN_SECTION. Rerunning the calculation will add trajectories to
the swarm.

24 Chapter 4 Exercise 3: vMCG

Exercise 4: DD-vMCG 5
„ Sometimes it is best to make a tea or coffee
and come back to it.

— Graham Worth
(Professor of Computational Chemistry)

In this exercise we will look at how to set up and run a DD-vMCG calculation
in which the potential energy surfaces are generated on-the-fly using an
external quantum chemistry code. The ideas behind the calculations are the
same as those used in the vMCG and MCTDH calculations of the previous
exercises. Calculations are run using normal mode coordinates as used in
the vibronic coupling model.

5.1 Butatriene Cation using Molpro

Get the input file but_dd.inp and the directory but_dddata. Before run-
ning a calculation, you need to

• Obtain an optimised ground-state state and frequencies. These define

the coordinates of the system.
• Decide on the level of theory for the excited states.
• Know how to write the input for an excited state calculation, including
gradients, non-adiabatic couplings and Hessian.

In this calculation of the butatriene cation we will use the Molpro program,
with 2 cation states calculated with a CAS(5,6) calculation.

In but_dddata there is a file but_fc_cas_freq.out which is a CAS(6,6)

calculation of the neutral ground-state frequencies. Note the keyword
freq=hpmodes which is needed to output the frequencies at higher precision
than the standard output.

25
 There is also a file template.dat which is a template for the quantum chem-
istry calculations. It is like a Molpro input file, but certain keywords which
are not always needed, such as that to calculate the Hessian, are bracketed
with markup-like keywords. The space where the geometry and molecular
orbitals are written is defined with <geometry> and <mo-coefficients>.

Finally, there is a checkpoint file start.chk containing the molecular

orbitals for the CAS calculation at the Frank-Condon point.

The Quantics input file but_dd.inp is different from the MCTDH style of
input. In the RUN_SECTION the keywords direct = nmodes specifies a
direct dynamics calculation in normal modes. ngwp = 5 species a basis of
5 GWPs.

The INITIAL-GEOMETRY-SECTION replaces the PRIMITVE-BASIS-SECTION

and INIT_WF-SECTION sections. It specifies the atoms in the molecule,
along with the Cartesian coordinates of the initial (Frank-Condon) geometry.
It also specifies the labels and frequencies of the normal modes. Other
information such as the number of electronic states (nstates=2) and the
initial state (init_state = 2)

Note that many of the normal mods have the keyword freeze at the end
of the line. This modes are frozen and only the 5AU and 14AG modes are
included in the dynamics. This 2D model is the same as that of the MCTDH
exercises.

The final new section is the DIRDYN-SECTION. This has the information
required to run the direct dynamics interface. The main keywords that need
to be set are:

• data = but_dddata defines the directory with the DD files and where
the database will be stored.
• transfile = but_fc_cas_freq.log defines the file with frequencies
• qcprogram = molpro defines that Molpro will be used
• method = cas defines that CAS will be the electronic structure method
used.
• ener0 = -152.9326464722 defines the energy of the ground-state
equilibrium geometry (in Hartree)
• dd_diab=global defines the diabatisation method (propagation dia-
batisation)

26 Chapter 5 Exercise 4: DD-vMCG

 • dbmin = 0.20 defines how far the molecule moves before a new QC
calculation is performed.
• db = rdwr A read-write database is used. I.e. new calculations are
run and added to the DB when required.
• nbasis = 44 number of Molecular orbitals
• subcmd = /user/gworth/bin/run_molpro15 script used to run the
Gaussian calculation

Run the calculation

• mkdir but_dd
• quantics but_dd

Much of the analysis follows that used for the vMCG and MCTDH calcula-
tions. The diabatic state populations can be plotted using statepop and
the potential using showsys. To obtain the adiabatic populations it is not
possible to do the full transformation as this requires multi-dimensional
integrals. Instead, we use a simple “saddle-point” approximation that trans-
forms the basis only at the centre of the GWPs. In the results directory
use

• adpop -saddle0

and then plot the results with gnuplot.

The centres of the GWPS can be plotted using gwptraj. In addition, we can
now plot the molecular structure along the trajectory using the program

• ddtraj

In response to the question Enter: Trajectory No. (Max: 5) enter

a number between 1 and 5. A window is opened using the Jmol program.
Select

• Tools −→ Animate −→ Once

to watch the movie. It is possible to watch the average structure etc.

Finally, the database can be examined to see the structure sampled. In the
directory but_dddata, there is a file database.sql. This is an sqlite file
containing all the structures, energies, gradients, etc. from the quantum

5.1 Butatriene Cation using Molpro 27

 chemistry calculations. This data forms the potential surfaces used in the
dynamics. Type

• checkdb

This will report the information in the database and how many structures.
To view the structures type

• rddddb
• jmol plot.db

where the second line starts a jmol window to plot the structures extracted
from the database using the rddddb program.

The results so far were obtained with a small basis (5 GWPs) and the DB
only contains 9 points. To improve the quality of the data, edit the input file
and increase the number of GWPs to 20. Now rerun the calculation using

• quantics -w but_dd

The -w options means that the output in the but_dd directory will be
overwritten, but the data already collected in the but_dddata database will
be added to.

At the end of this calculation re-run the analysis.

5.2 Water Excitation using Molpro

The second example is a calculation of the photo-dissociation of water

using Molpro. Take the input file h2o-4s-dd.inp and the directory for the
database h2o-4s-dd_data

As before, the DB directory contains a file with the ground-state frequencies

water-opt-nosym.out along with a template.dat file. This time the tem-
plate is that for a Molpro input file for a SA4-CAS(6,8)/6-31G* calculation.
The file init-orbital.orb contains the CASSCF molecular orbitals at the
Franck-Condon point.

In the Quantics input file h2o-4s-dd.inp the sections

28 Chapter 5 Exercise 4: DD-vMCG

HAMILTONIAN-SECTION_system
modes | 1A1 | 2A1 | 3B2
1.0 |2 capa
end-hamiltonian-section

LABELS-SECTION
capa=CAP[10.0,0.01,3,1]
end-labels-section

define a CAP along the 2A1 vibration, which is the anti-symmetric O–H
stretch.

Run the calculation and analyse the results.

5.2 Water Excitation using Molpro 29

Exercise 5: NOCl 6
Photodissociation

6.1 The Absorption Spectrum

The photodissociation of NOCl is a simple photo-chemical reaction. After

excitation from the ground to the first excited state, S0 → S1 , the chlorine
atom dissociates on a femto-second time-scale. This results in a broad band
for the absorption spectrum.

The calculation consists of two stages. First, the ground state wavefunction
is generated by energy relaxation of an initial guess wavefunction on the
ground state surface, S0 . The second stage then places this wavepacket on
the excited state surface, S1 , leading to photodissociation.

Copy the files nocl0.inp, nocl0.op, nocl1.inp and nocl1.op to your

directory, and create there the directories nocl0 and nocl1.

To find the ground-state eigenfunction, we first perform a relaxation calcu-

lation. Type

• quantics nocl0

To perform the photo-dissociation calculation, the wavepacket from the

nocl0 calculation is read in and projected on the S1 surface. Type

• quantics nocl1

This calculation takes a few minutes.

The calculation can now be analysed. Move to the directory nocl1 which
contains all the data files from the propagation. To watch the system
dissociating, start the showd1d program

• showd1d -inter

31
 100
90
80
70
60
50
40
30
20
10
0
-10
0.6 0.8 1 1.2 1.4 1.6 1.8 2
Energy[eV]

Fig. 6.1: The absorption spectrum for the NOCl molecule on excitation to the S1
state.

and follow the menu options to plot. First type 1. This gives a 3D plot of
the wavepacket as a function of time. Next type 10 to change the plot task
and choose 2 to plot time slices. Enter 1 to plot these. You can now make a
movie by typing 280 and selecting 1 or 2. With option 1 you need to press
return to go through the frames.

To plot the spectrum, type

• autospec -inter

You now need to enter the required range for the spectrum, a damping time
and a damping order. For the range enter 0.6 2.0 ev. For the damping time
0, for the damping order 1. To plot the spectrum type 1. The result is shown
in Fig. 6.1. Note that the spectrum shown is the Fourier-transform of the
autocorrelation function times the energy. Hence it is assumed that the
ground state energy is at zero, such that energy equals excitation energy. If
this is not the case, use option 802 to shift the energy scale.

Save the spectrum to a gnuplot file using option 4 = save plot to a gnuplot
file (use after 1 or 2). Give the name spectrum1.pl. This file contains the
data and commands to plot using gnuplot spectrum1.pl

• gnuplot spectrum1.pl

The program showsys is a powerful tool for plotting 1D and in particular

2D views on wavepackets and potentials. Type

• showsys

32 Chapter 6 Exercise 5: NOCl Photodissociation

 2.6

2.4

2.2
rv [au]

1.8

3.8 4 4.2 4.4 4.6 4.8 5 5.2 5.4 5.6

rd [au]

Fig. 6.2: Overlay plot, wavepacket on potential. The Wavepacket density is shown
for the times t = 0, 10, 20, 30 fs. The density is obtained by integrating
|Ψ|2 over all angles, whereas the potential contour lines are obtained by
fixing the angle to 2.1 rad.

A menu appears which allows various options to be set. Go to menu point

10 (type 10), and change the plot task to 2 = plot pes (type 2). Next input a
1 three times and a 2D cut through the surface (with theta fixed to 1.545
radians) will pop up. Now use menu point 20 = change coordinate section,
i.e. chose another cut. If one specifies only 1 coordinate with an x and the
others with numbers, a 1D plot will appear. After you have played around
enough, go back to menu point 20 and input

rd x
rv y
theta 2.1

Then use menu point 5. You will be asked for a file name. Chose any
convenient name, e. g. xyz. The plot data is then written to the file xyz for
later use.

Next we want to inspect the wavefunction. Go to menu point 10 and chose

5 = plot reduced density. The density, i.e |Ψ|2 integrated over all coordinates,
except those specified by x and y (that is integrated over all angles in the
present case), will be shown. Input a 1 three times and you will see the
initial density. Pressing RETURN will display the density propagated by one
time step, and so on. After you have returned to the menu, chose point

6.1 The Absorption Spectrum 33

 400 = Overlay plots and then 410 = File for overlay and enter the file name
(xyz). Turn off the key using option 240. After inputting 1’s you will now
see an overlay plot, i. e. the wavepacket on top of the contour lines of the
potential.

With menu points 240 and 245 one may switch off the legend (or keys) and
the title. Menu point 285 allows to take larger time steps and with point
280 one may switch to different plot forms, e. g. to plot all time slices at
once. Such a plot is shown in Fig. 6.2.

Inspect the Ascii files of the output directory, in particular output, log, and
timing. The file input contains a copy of the input file, the options, and the
operator file. Thus, it tells you exactly what you have been doing. Since
an NOCl run is so fast, NOCl is ideally suited for testing. Just play around
with it! You may e. g. change the numbers of single particle functions or
alter the integrator accuracies.

6.2 Computing excitation and reaction

probabilities

The analysis of scattering processes, i. e. the computation of excitation

and reaction probabilities, is performed by evaluating the quantum flux
going into a particular channel. The quantum flux is determined by the
interaction of the time-dependent wavepacket with a CAP. The program
flux performs the necessary analysis. The psi file is read and the energy
resolved flux is computed and written to the flux file. Additionally this flux
is divided by the energy distribution of the initial wavepacket to obtain
the transition or reaction probabilities. The latter step requires in general
that the quantics program has computed the energy distribution of the
initial wavepacket and has written it to the enerd file. Besides the flux file,
flux creates the files flux.log, gtau and wtt. When the file gtau is present
the flux program skips the time consuming evaluation of the integrals
⟨Ψ(t)|W |Ψ(t + τ )⟩ but reads the function g(τ ) from the gtau file. (See the
MCTDH review, section 8.6.3, Eq.(199), for more details). This allows to
re–do the Fourier integrals for another energy interval very quickly. The
option -w enforces the re–calculation of g(τ ).

The file wtt contains the expectation values Wtt = ⟨Ψ(t)|W |Ψ(t)⟩. This in-
formation is useful for checking that the absorption process has finished.

34 Chapter 6 Exercise 5: NOCl Photodissociation

The shell script plwtt visualises the function Wtt , while plflux and plflux -r
visualise the flux and the transition probability, respectively.

The program flux cannot only determine the total energy resolved flux going
into a particular arrangement channel (i. e. going into a particular CAP) but
can also determine the flux which is projected onto final quantum states
or which is weighted by an operator. This is probably best demonstrated
by an example. We also want to plot the flux as a function of time so want
to calculate the expectation value of the flux operator in the dissociation
channel.

To analyse the result, move into the directory nocl1. First plot the flux as a
function of time. This is contained in the file expectation. Start up gnuplot
and type

• plot "expectation" u 1:2 w l

This plots the norm which is close to 0 at the end of the propagation.

• plot "expectation" u 1:3 w l

plots the flux. There is a set of peaks, dominated by a first, fast component
and a second, slower component.

Next evaluate the total flux as a function of energy and save it in the file
flux.0

• flux -w 0.61 2.0 ev rd

• mv flux flux.0

To display it, type

• plflux -f flux.0

As you will notice, the plot looks very similar to the absorption spectrum
calculated above (and shown in figure 6.1). It is not identical, though, as
the definition of an absorption spectrum contains a factor ω, the energy of
the absorbed photon. This multiplication is omitted when autospec is run
with the -FT option. Type

• autospec -FT 0.61 2.0 ev 0 1

6.2 Computing excitation and reaction probabilities 35

 which writes the spectrum to the file spectrum.pl. Now the spectra are
identical as one observes when typing

• plflux -G -f flux.0 -d spectrum.pl

This shows the flux and the Fourier transform of the autocorrelation function
on top of each other.

Next we modify the flux by letting the operator of the vibrational energy act
on it.

• flux -w -O vib -u 200. 0.61 2.0 ev rd

• mv flux flux.op

To view this

• plflux -G -f flux.0 -d flux.op

and you will see the total flux in comparison with the vibrational energy
weighted flux. It is now clear that the structures in the spectrum are due
to vibrational excitation of the NO fragment. By the way, via the option
-u 200 the modified flux was multiplied by the factor 200. This was done
to make it comparable with the total flux. Usually the option -u is followed
by an energy keyword, e.g. -u ev, to transform the weighted flux from a.u.
to a desired energy unit.

One may divide the weighted flux by the total flux to observe the vibrational
energy content by using the option -ed flux.0

• flux -ed flux.0 -O vib -u ev 0.61 2.0 ev rd

• mv flux flux.op_r

To visualise the quotient weighted-flux/total-flux, which is related to the

reaction prabability, type

• plflux -G -r -f flux.op_r

The structures below 0.8 eV are numerical noise because one divides a very
small number by another very small number. Obviously, the higher the
energy of the absorbed photon, the larger is the vibrational energy of the
NO fragment and the higher the chance of dissociation.

36 Chapter 6 Exercise 5: NOCl Photodissociation

100
v=0
v=1
v=2

80

60

40

20

0
0.6 0.8 1 1.2 1.4 1.6 1.8 2
Energy[eV]

Fig. 6.3: The total (full line) and projected flux of dissociating NOCl. The projec-
tion is on the vibrational states of the NO fragment, v=0, v=1, and v=2,
respectively.

Finally we demonstrate the use of projectors. To project onto the lowest 3

vibrational eigenfunctions of the NO Hamiltonian, type

• flux -w -P 2 eigenf vib 1 % 0.61 2.0 ev rd

• mv flux flux.1
• flux -w -P 2 eigenf vib 2 % 0.61 2.0 ev rd
• mv flux flux.2
• flux -w -P 2 eigenf vib 3 % 0.61 2.0 ev rd
• mv flux flux.3

To display the results type

• plflux -G -f flux.0 -d flux.1 -e flux.2

• plflux -G -a 1.0 -y 25. -f flux.0 -d flux.2 -e flux.3

The first plot shows the flux projected onto the vibrational ground state and
the first excited state, respectively, in comparison with the total flux. The
second plot is similar, but shows the flux projected onto the first and second
excited state, respectively. The Fig. 6.3 displays the total and projected
projected flux, similarly to the plots generated above.

6.2 Computing excitation and reaction probabilities 37

Bibliography

[1] J. Coonjobeeharry, K. E. Spinlove, C. Sanz Sanz, et al. „Mixed-quantum-

classical or fully-quantized dynamics? A unified code to compare meth-
ods“. In: Phil. Trans. Roy. Soc. A 380 (2022), p. 20200386 (cit. on p. 22).
[2] A. Ferretti, G. Granucci, A. Lami, M. Persico, and G. Villani. „Quantum
mechanical and semiclassical dynamics at a conical intersection“. In:
J. Chem. Phys. 104 (1996), pp. 5517–5527 (cit. on p. 22).
[3] J. C. Tully. „Molecular dynamics with electronic transitions“. In: J.
Chem. Phys 93 (1990), pp. 1061–1071 (cit. on p. 9).

39
Glossary

CAP

Complex absorbing potential used to absorb the wavepacket. 10

DD-vMCG

The direct dynamics version of variational multiconfigurational Gaus-

sian method whereby potential energy surfaces are calculated on-the-
fly (as required) utilising external Quantum Chemistry software. 2

DVR

Discrete variable representation. 16

GWP

Gaussian wavepacket. 21

HAMILTONIAN-SECTION

In an operator file this is the section where one defines the operator. 3

INIT_WF-SECTION

In an input file this is the section where the initial wavefunction is

defined. 4, 7

41
 input file

This is the input file for the calculations. Here we link to an operator
file (where appropriate), the basis set, basis functions, the initial
wavefunction and the run parameters of the calculation are set. 4

MCTDH

Multi-Configuration Time Dependant Hartree (method). 21

OP_DEFINE-SECTION

In an operator file this the section that defines the file - including the
title. 3

operator file

This is where the operators for a calculation are defined. Here the op-
erator file is defined with sections for the Hamiltonian and parameters.
3

OPERATOR-SECTION

In an input file this is the section that defines the operator file being
used. 4

PARAMETER-SECTION

In an operator file this is the section where one defines the parameters
of the operator. 4

PRIMITIVE-BASIS-SECTION

In an input file this is the section where the “primitive basis set”, or
the grid, is defined. 4, 21

42 Glossary
adpop

Results/Analysis: Calculates the diabatic and adiabatic populations

of the electronic states. The calculation of one and two dimensional
adiabatic densities is also possible.. 17

analysis

Results/Analysis: A driver for calling the appropriate analysis pro-

grams. 20

autospec

Results/Analysis: Computation of a spectrum by Fourier-transforming

the autocorrelation function. 19

gwptraj

Results/Analysis: Plots the centres of GWP basis functions in vMCG

(and G-MCTDH) calculations.. 22

-inter

Results/Analysis: “inter” is an abbreviation of “interactive” and hence

is the command that calls for the interactive menu for results/analysis
programs. 17, 18

rdauto

Results/Analysis: Reads the auto steps from file "auto" and enables
plotting of the autocorrelation function.. 18

rdcheck

Results/Analysis: For plotting the trajectory of the centre of the

wavepacket, UPDATE THIS. 6, 16

showsys

Results/Analysis: For plotting potential energy surfaces, with a nu-

merical selection based interface. 5, 15, 17

Glossary 43
 statepop

Results/Analysis: Outputs the diabatic state populations.. 22

RUN-SECTION

In the input file this is the section which defines the calculation such
as the propagation time.. 4

SPF

“Single-particle” basis function. 16, 21

SPF-BASIS-SECTION

In an input file this is the section where the number of “single-particle”

basis functions will be defined. 4

Tully Models

A set of three model 1-dimensional potential energy surfaces set up

by John C. Tully to benchmark scattering calculations incuding non-
adiabatic effects (Tully JCP (90) 93: 1061). 9

vMCG

The variational multiconfigurational Gaussian method where the MCTDH

SPFs are replaced with variational Gaussians. 2, 21

44 Glossary
Glossary 45