This document will provide a hands-on tutorials demonstrating the ADF-GUI, BAND-GUI and so on. All of these (ADF-GUI, BAND-GUI, DFTB-GUI, ReaxFF-GUI and so on) are very similar and share a lot of options. So collectively these will just be called the GUI.
The first five tutorials provide an overview over all GUI modules, using ADF as computational engine. If you are new to the GUI, we advise you to study the tutorials in the order presented.
The advanced ADF-GUI tutorials focus on particular features of the ADF-GUI, or on some particular application.
Next there are some specialized tutorials that show you how to use the GUI with BAND, DFTB, MOPAC or ReaxFF.
We estimate that you will need about half an hour for each part of the tutorial, for reading and performing the hands-on activities. Typically the calculations should give results that are very close to the results shown in the tutorials. However, in some cases minor changes may occur depending on hardware and software versions.
Many tutorials are also available as videos. The current videos are actually the tutorial videos for ADF2009, so there will be differences with the current tutorials.
Additional information may be found in the GUI manual
We will assume that you have installed the full ADF package on your machines.
The most convenient way to start the GUI is to first start ADFjobs, one of the GUI modules. Using the SCM menu in ADFjobs (either the text SCM, or the SCM logo) you can next start any other GUI module that you have a license for.
Depending on your platform, you start ADFjobs in a slightly different way:
In this tutorial we will assume that you have added $ADFBIN to your PATH. We advise you to do this if you have not already done so. If you do not wish to add $ADFBIN to your PATH, you must start the utilities using
$ADFBIN/adfjobs &
instead of just
adfjobs &
On all UNIX platforms, including Mac OS X, the GUI modules are XWindow (X11) programs. Your environment should have been set up correctly to run such programs. In particular, you need to make sure that your DISPLAY environment variable has been set properly.
You can test this easily by starting a simple X program from your shell:
xcalc &
Assuming the xcalc program is in your $PATH, a small window should appear with the image of a calculator. If not, please consult your local system administrator for additional help. Typical problems have to do with your DISPLAY variable, and with security issues when trying to connect to the X server.
In this tutorial we will assume that you have installed the ADF package using the Installer with all the options set to default.
You can start the GUI (adfjobs) by double-clicking its icon on the desktop.
Macintosh users have several easy ways of starting the GUI modules: just like on any other UNIX system (thus, using the command line), or using the ADFLaunch program.
The easiest way is using the 'ADFLaunch' program. When you start it, ADFjobs will be started automatically. Next you can use the SCM menu to start the other ADF-GUI modules.
The ADF-GUI is the Graphical User Interface for the ADF package. It consists of several modules for specific tasks. Those modules work together and exchange information. You should make sure you have no firewall blocking TCP/IP communications inside your machine.
All of the ADF-GUI (and BAND-GUI etc) modules have one common SCM menu on the top left of the window. On most systems it is represented by a small SCM logo. You can use the commands inside the SCM menu to start other GUI modules (or switch to them).
In general when selecting a GUI module from the SCM menu it will start and open the current job. If that module is already open with the current job, it will be activated (brought to the foreground). The current job is the selected job in ADFjobs, or the job open in some other GUI module if you use the SCM menu in that module.
The most important exception is opening an Input module (like ADFinput, BANDinput, ...) or ADFcrs (COSMO-RS) in ADFjobs. In that case the selected job will be ignored, and you can start working on a new calculation. To open the selected job in ADFinput, you need to click the icon in front of the job.
SCM → Preferences
GUIprefs ($ADFBIN/guiprefs) allows you to adjust and save numerous GUI preferences, such as color schemes, environmental variables, etc. The preferences will be used by all ADF-GUI modules.
SCM → ADF Input
SCM → BAND Input
SCM → DFTB Input
SCM → UFF Input
SCM → ReaxFF Input
ADFinput ($ADFBIN/adfinput) helps users to easily create ADF jobs. You can use ADFinput to define your molecule (geometry), pre-optimize it, and to set details of your ADF job using an easy-to-use graphical user interface. ADFinput will generate the basic job script for you. This script takes care of running ADF and property programs as required.
The same module can actually create jobs using different methods: ADF, BAND, DFTB, MM, MOPAC, Open Babel, QMMM, QUILD, ReaxFF and UFF. After starting it, you can simply change the method to use without starting a different module. Alternatively, if you use the BAND Input menu command the same module will start up, but will immediately start in BAND-mode (to set up a BAND calculation). Depending on your license, not all options will be available.
SCM → COSMO-RS
ADFcrs ($ADFBIN/adfcrs) enables ADF users to easily select compounds, create COSMO-RS jobs, run the jobs, and visualize the results.
SCM → View
ADFview ($ADFBIN/adfview) displays volume data, such as electron densities, orbitals, electrostatic potentials and more. You can also use it to visualize scalar atomic data like charges, some tensor data, and AIM (Bader) results.
SCM → Movie
ADFmovie ($ADFBIN/adfmovie) follows geometry steps as performed by ADF during geometry optimizations, IRC calculations, etc. It can be used during the calculation to monitor the progress (based on information from the logfile), or it can be used to analyze the geometry changes after a calculation. It is also used to display normal modes calculated with a frequency calculation.
SCM → Levels
ADFlevels ($ADFBIN/adflevels) generates a diagram showing the energy levels of a finished calculation. You can interact with it: show an interaction diagram (how the molecular orbitals are constructed from fragment orbitals), show labels, occupations, orbitals, etc.
SCM → Logfile
ADFtail ($ADFBIN/adftail) shows the contents of a text file, updating when the text file grows (like the UNIX tail -f command). It is typically used to monitor the 'logfile'. The progress of an ADF calculation is always written to this file.
SCM → Output
ADFoutput ($ADFBIN/adfoutput) shows the output of ADF (or any other text file). It will analyze the output and provide quick links to sections of interest.
SCM → Spectra
ADFspectra ($ADFBIN/adfspectra) shows spectra calculated by ADF. It can show IR, Raman, excitation and CD spectra, as well as a DOS plot. For some spectra it can also perform additional tasks (using other ADF-GUI modules), like displaying normal modes or orbitals.
SCM → Band Structure
Bandstructure ($ADFBIN/bandstructure) shows dispersion spectra like the band structure of solids, or phonon spectra, as calculated by for example Band or DFTB.
SCM → Dos
ADFdos ($ADFBIN/adfdos) shows DOS-like results. You can easily select which partial DOS to show by selecting atoms, and you can even select to show the GPDOS for select atoms and L-shells.
SCM → Jobs
This utility ($ADFBIN/adfjobs) manages your ADF jobs: run a job on your local machine or on remote machines. It also serves as a interface to all files belonging to your job, and it serves as a convenient launcher of the other ADF-GUI modules.
Some keyboard shortcuts for menu commands used in the tutorials are platform specific. For example:
In general, on Linux and Windows machines use the ctrl key to access these shortcuts for the menu commands, on a Mac use the cmd key. In the tutorials, these command keys may be mentioned as for example cmd/ctrl-S
If you are using a Mac, the GUI uses the X11 program to run. The X11 program itself may intercept the menu-shortcuts like Cmd-C, Cmd-H, and so on. You can change this using the Preferences command from the X11 menu (Enable key equivalents under X11 should NOT be checked).
A three-button mouse is also very convenient for using the ADF-GUI, and on a Mac you can use a Magic mouse for this purpose. To get three buttons (instead of the standard one or two), download and install the free utility BetterTouchTool, and configure it to add a middle click.
This tutorial will help you to:
On a Unix-like system, enter the following command:
cd adfjobs &
On Windows, one can start ADFjobs by double-clicking on the ADF-GUI icon on the Desktop:
double click the ADF-GUI icon on the Desktop
On Macintosh, use the ADFLaunch program to start ADFjobs:
Double click on the ADFLaunch icon
Make sure X11 allows you to use keyboard shortcuts (you only need to do this once):
Press Cmd-,
If nothing happens everything is fine.
If the X11 Preferences dialog appears:
uncheck the "Enable key equivalents under X11" check box and close that dialog.
Note that the directory in which ADFjobs depends on how you start ADFjobs, so your screen might look different.
We prefer to run the tutorial in a new, clean, directory. That way we will not interfere with other projects. ADFjobs not only manages your jobs, but also has some file management options. In this case we use ADFjobs to make the new directory:
Select the File → New Directory command (thus, the New Directory command from the File menu) Rename the new directory by typing 'Tutorial' and a Return
Change into that directory by clicking once on the folder icon in front of it
Now we will start ADFinput in this directory using the SCM menu:
Select the SCM → ADF Input menu command:
use the SCM menu on the top left (the SCM logo on most platforms),
next select the ADF Input command from the menu.
The ADFinput module should start:
The ADFinput window consists of the following main parts:
ADFinput has an Undo command, which works on your molecule (thus not on your input options).
If you make a mistake while making changes to your molecule, just use the Edit → Undo menu command to go back in time. You can Undo more than one step, or Redo a step (with Edit → Redo) if you wish to do so.
The molecule we are going to create is ethanol.
First we will draw the two carbon atoms, next the oxygen atom, and after that we will add all hydrogen atoms at once. Finally, we will pre-optimize the geometry within ADFinput.
To create an atom, you need to select an atom tool.
Select the C-tool by clicking on the button with the 'C'
Back glow is added to the 'C' button to indicate that you are using the C-tool. Also, the status field in the left bottom corner shows 'C tool, create single bond' to indicate that you are using the C-tool.
Now create the first carbon atom:
Click somewhere in the drawing area
One carbon atom has been created.
Note that:
Click somewhere in the drawing area to create the second carbon atom
A second carbon atom has been created, bonded to the first atom.
The atom will be created along the 'bonding line', at a distance that corresponds to a normal C-C single bond distance. That is, the bond length is constrained while drawing.
The newly created atom becomes the new selection, and you are still in bonding mode. The next bond will be created to the carbon atom just created. And you are still using the C-tool.
To create the oxygen atom you need to switch to the O-tool:
Select the O-tool by clicking on the button with the 'O'
With the O-tool, create an oxygen atom bonded to the second carbon;
Click somewhere in the drawing area
The oxygen atom has been added.
For now, we are done using atom tools, so go back to the select tool:
Select the select-tool by clicking on the button with the arrow (or press the Esc key)
Now many hydrogen atoms need to be added. You can do this using the H-tool, but a much easier method is to use the Atoms → Add Hydrogen menu command:
The 'Add Hydrogen' menu command works on the selection only, when present. Thus, only one hydrogen atom would be added to the oxygen atom. This is not what you want. So first we make sure that nothing is selected by clicking in empty space.
Click in empty (drawing) space
Now no atoms are selected any more.
Select the Atoms → Add Hydrogen command
Many menu commands have shortcuts. In this case you can also use the shortcut (ctrl-H or cmd-H, depending on your platform) as an alternative. The shortcuts are indicated in the menu commands.
All atoms will be saturated with hydrogen atoms. And you have created an ethanol molecule, though the geometry is still far from perfect.
Now use the optimizer that comes with ADFinput to pre-optimize the geometry.
Click on the pre-optimizer button (the rightmost 'cog wheel' button of the menu bar)
The geometry of the molecule will be pre-optimized, using UFF by default. Note you can select another pre-optimizer via the Preferences, or use a different pre-optimizer by right-clicking on the cog wheel and selecting the method to use from the pop-up menu.
In the status field below the drawing area you can follow the pre-optimization iteration number and the energy relatively to the starting configuration.
You can rotate, translate, and zoom your molecule using the mouse.
You need to drag with the mouse: press a mouse button, and while holding it down move it. Which mouse button, and which modifier key you press at the same time, determines what will happen:
| Rotate | Left |
| Rotate in-plane | ctrl-Left |
| Translate | Middle, or Alt-Left |
| Zoom | Mouse wheel, Right, or cmd-Left (drag up or down) |
The rotate, translate, and zoom operations change how you look at the molecule, they do not change the coordinates.
Click once somewhere in empty space to make sure nothing is selected. Click with the left mouse button, and drag: your molecule will rotate. Click with the left mouse button with the ctrl-key, and drag: your molecule will rotate in-plane. Click with the middle mouse button (or left with alt-key), and drag: your molecule will be translated. Click with the right mouse button (or left with cmd key), and drag up and down: you will zoom closer to or away from your molecule. Use the mouse wheel, if you have one: you will zoom closer to or away from your molecule. Using all these options, try to position the ethanol as closely as possible as in the following image:
The atoms of your molecule will have a pre-defined color, as visible in the molecule editor. You can change these colors in two ways: either change the global default color per element, or override the default color for a specific atom. The first method uses the Preferences command, the second method uses the 'Atom Details panel' via the Atoms → Details (Color, Radius, Mass, ...). In this tutorial we will change the default color of all Carbon atoms using the Preferences command:
Use the SCM → Preferences menu command to start the preferences module
In the preferences window, go to the 'Atom Colors' panel (use the panel bar Color menu)
Click on the '+' button, and select C (Carbon) from the periodic table that pops up.
You can now change the color for the carbon atoms either by clicking on the colored block next to the C label
Click on the colored C-block Select your favorite color, and dismiss the Color dialog (click the OK button on most platforms) Click the Preview button in the bottom of the GUIprefs panel.
In ADFinput the colors of the carbon atoms will now be adjusted to the color that you have selected.
The 'Preview' button shows the effect of your new preferences in all open GUI modules, but without saving them. Once you are content with your changes, you can use the 'Save' button to store the new defaults.
If you close the Preferences window without saving, your original settings will be restored in the open modules.
To revert to the default value of carbon, you can remove the carbon color in Preferences window by pressing the - button in front of the C color line:
Click the '-' button in front of the C color line Click the 'Preview' button
In ADFinput, your carbon atoms now should have their default color again (dark gray).
Use the File → Quit menu command to close the GUIprefs module (the Preferences window)
When quitting the Preferences module without saving, any changes made visible via the Preview button will be undone.
You can save a picture of your molecule using the 'Save Picture ...' command from the File menu.
The format used is the PNG format. You can change this using the File → Picture Format menu command. You can also change the resolution. A smaller resolution will result in a smaller file, but will reduce the quality.
Via the Preferences command it is possible to save your preferred format.
Select the File → Save Picture ... command Enter the name for your picture: ethanol Click the 'Save' button
A picture will be saved to disk containing the image of your molecule. Only the drawing area is saved in the picture, not all the input options.
Rotate your molecule into the following position:
Click once on the top carbon atom Use the Select → Select Connected menu command
As you will notice, all atoms directly connected to the selected atom are added to the selection. Alternatively, you can also make a selection by shift-clicking on the elements you want to select.
Click in empty space Click on the top carbon atom Shift-Click once (without moving) on each of the top hydrogen atoms
This has almost the same effect (in this case you have not selected the second carbon atom).
We are now trying to make an eclipsed geometry.
ctrl-Click with the left mouse button in one of the selected hydrogens, and drag around to rotate the selection in-plane. Rotate the hydrogen atoms in an almost eclipsed position.
You can move the selection by clicking in a selected object, and dragging with the mouse. All usual operations are possible: rotate, rotate in-plane, translate and zoom. Zoom in this case means moving the selection perpendicular to the screen.
In the status field you see the current rotation angle.
You have to click and start dragging at a selected item. If you click and drag in space you will move the entire molecule.
Click in empty space to clear the selection Click on the pre-optimize button
The optimizer will bring the structure back to the original staggered geometry. If it does not complete, repeat this step until it does.
First select the oxygen atom and the connected hydrogen atom.
This time we make the selection by dragging a rectangle around all objects that we want to select.
Using the left mouse button together with the shift key, drag a rectangle around the oxygen and hydrogen atom.
Release the mouse button (and the shift key)
The oxygen atom and the hydrogen atom are selected.
In the status area you see the distance between the selected atoms, information about the bond, and a slider.
You can set the distance to any value you wish by editing it, or (most conveniently) by using the slider.
Normally the smallest group of atoms will move to create the requested distance. If you press the control key while using the slider the last atom selected will be in the group of atoms to move.
Use the slider to move the H atom
Select first one of the top hydrogens by clicking on it Next, extend the selection (shift key) by clicking on the top carbon atom Finally, extend the selection (shift key) by clicking on another top hydrogen atom
In the status area information about the bond angle of the selected three atoms is given, and the slider is again visible. You can change this value to a value you like, most conveniently using the slider.
By selecting four atoms we get information about the dihedral angle. And of course you can also change it, again most conveniently using the slider.
Move the molecule such that you can see all atoms Select first one of the top hydrogens by clicking on it Next, extend the selection (shift key) by clicking on the top carbon atom Next, extend the selection (shift key) by clicking on the next carbon atom Finally, extend the selection (shift key) by clicking on the oxygen atom
To view the coordinates we have to go to a different input panel. The input panels can be selected using the panel bar on the top of the input panels, the right half of the window.
In the right side of the ADFinput window: Click on the Model tab in the panel bar Select the Coordinates command
You get a list of all Cartesian coordinates. They will be updated in real time when you make changes to the molecule, and you can also edit the values yourself. In that case, the picture of the molecule will be updated automatically.
Note that some atoms are highlighted. These are the currently selected atoms.
The 'Move Atom(s)' buttons will move the selected atoms up or down. In this way you can re-order the atoms.
You may, in some cases, wish to use Internal coordinates. These will be updated in real time as well, and you can also edit the values yourself. If you re-order the atoms while using the Internal coordinates, the Z-matrix will be recalculated from scratch. The Z-matrix panel is in the Details tab, but we will use the search option to get to it:
Click the search icon at the right of the panel bar Enter the text "z-ma" in the search field
Note that in the search results the Z-Matrix panel is highlighted, and note that the search is not case sensitive.
Click on the selected option (Z-Matrix), or use the return key
Before making some changes, let's re-optimize. We first select the 'Main' panel so the coordinates will not be visible during the pre-optimization. Otherwise this may slow down the pre-optimization.
Click on the "Main" tab Click on the pre-optimize button
Let's try to change the CH2OH group in a COOH group.
Thus, we need to:
After this, we will revert to the ethanol molecule.
First: delete one hydrogen
Click in empty space to clear the selection Click once on the hydrogen to delete, it will be selected
Press the backspace key
The selected atom is removed.
Next, we will change a hydrogen into an oxygen atom
Select the O-tool (or press the 'O' key) Double-click on the hydrogen that should change into an oxygen
Now the oxygen atom is singly bonded to the carbon, we need to change this into a double bond.
Click on the bond between the carbon and the new oxygen Use the Bonds → Bond Order → Double menu command (or just press the '2' key)
The single bond has changed into a double bond.
Another way to modify a bond type is to click on the bond once which will select this bond. Then click on the bond tool in the menu bar ('ball and stick' logo to the right of the start), and select the proper bond type.
The fastest way to change the bond order is to use the keyboard shortcuts. If you have selected a bond, press 1 for a single bond, 2 for a double bond, 3 for a triple bond and 4 for an aromatic bond.
To get a reasonable geometry optimize the structure:
Click in empty space to deselect the bond Press the pre-optimize button If not converged, press pre-optimize again
Now, to revert to the ethanol molecule, we first remove the new doubly-bonded oxygen atom, and then add one hydrogen atom.
Click once on the doubly-bonded oxygen atom to select it Press the backspace key to delete it Select the H-tool (or press the 'H' key) Click once on the carbon atom connected to the oxygen
Note that this way you started bonding mode again, as indicated by the bond to the mouse position.
Click once in empty space to make a hydrogen atom (connected to the carbon!) Click once on the just created atom to stop bonding Repeat this to add a second hydrogen to the carbon atom Pre-optimize the molecule
Once again you have created an ethanol molecule.
You should always start by choosing the proper preset.
ADF has many different modes of operation. ADFinput presents the correct settings for some common calculations together as 'presets'.
So to optimize the geometry of the ethanol molecule we choose the proper preset:
Select the 'Geometry Optimization' preset from the 'Preset' menu.
Presets might change any number of input options. In this particular example, the main Task will change from 'Single Point' to 'Geometry Optimization'. You can easily see what fields have been updated by a preset: they are colored green.
If you would use this preset again later, the input values that are set by this preset will revert to the default values. If you made changes to those fields you will lose those changes.
This field, currently empty, has no special meaning for the calculation routine. It will be used as an identifier for your convenience in the result files of the ADF calculation. In this case, let's set the title to the name of the molecule:
Enter 'Ethanol' without quotes in the 'Title' field.
Note that as soon as you start typing in the title field, the color of the field changes to yellow. The yellow color here and in other cases indicates that you have made a change with respect to the default value for this field.
An important input option is the XC functional to use.
For this tutorial the default potential during the SCF is sufficient. So just leave this at the default value. For more accurate results you should select a better XC potential.
By selecting the LDA+GGA+METAGGA option for the XC energy ADF will print a table at the end of the calculation showing the energies with many different functionals:
Select the 'LDA+GGA+METAGGA' option for the XC energy after SCF
Again, since you made a change with respect to the default value the change is highlighted with a yellow color.
With the 'Basis Set' pull-down menu you select the basis set you want to use.
The menu gives access to the basis sets regularly used.
For this tutorial we will choose a very small basis set. This will yield less accurate results, but the calculation runs much faster. Obviously, if you want more accurate results you should use a better quality basis set. Thus:
Select 'SZ' from the 'Basis Set' pull-down menu
The ADF software package uses a numerical integration scheme for virtually everything it may calculate. The number specified here is roughly the number of significant digits in your results. If you leave the field blank ADF will select a reasonable default value.
Increasing the number makes the results more accurate, but will require substantially more computation time.
Similarly, decreasing the number will result in less accurate results, but you may get results faster. A too low integration accuracy often leads to convergence problems if the convergence criteria are too strict.
The default value will in most cases be fine, certainly for this tutorial.
The panels on the right side contain many more input options. You select a panel with the menus in the panel bar, or by searching for a particular option (as we did to locate the Z-Matrix panel). When searching for an option, any text in the panels will match, as well as from the help balloons. Also the corresponding ADF input keys will match.
The menu items use a color coding to show you which panels have been affected by a preset (green), by the user (yellow), or both (red).
As we will not do anything special right now, you do not need to change anything in other panels.
Finally you will want to save your input.
Select the File → Save command. Enter the name 'ethanol' in the Filename field.
Click on Save
ADFinput will now present a warning dialog:
So the requested option to print meta-GGA results after the final SCF will produce incorrect results when using frozen cores. Normally you would not accept this, but for the tutorial let's just continue with this set-up:
Click OK to dismiss the warning dialog
A new dialog pops up asking you if you really want to save the run script even though it will not be correct:
The run script will not be correct because we use the frozen core option together with metaGGA's, as we were told before. Accept this and save it anyhow:
Click Yes to save the run script
Now you have saved your current options and molecule information. The file will automatically get the extension '.adf'.
ADFinput has also created a corresponding script file. This script file has the same name, but with an extension '.run' instead of '.adf'.
In the ADFjobs module you can see what files have been created:
Click once in the ADFjobs window to activate it Click once on the triangle in front of the name of the job (ethanol)
You will see the .adf and .run files, and a .pid file that ADFjobs uses to store information. You might also see the picture that you saved, if you used the name 'ethanol' for it. Only the extensions are listed, so the real filenames are ethanol.adf, ethanol.run and ethanol.pid. Notice the job status icon (the open circle on the right) that ADFjobs uses to indicate a new job.
To actually perform the calculation (the geometry optimization of the ethanol molecule), use the Job → Run menu command in ADFjobs:
Make sure the ethanol job is selected in ADFjobs (it is if you followed the tutorial) Select the Job → Run command
This will execute the run script that has just been created. If you have never made changes in the ADFjobs setup, the default behavior is to run the job in the background on your local computer, using the Sequential queue. This queue will make sure that if you try to run more then one job at the same time, they will be run one after another.
Once your job starts running, ADFjobs will open ADFtail to monitor the progress of the calculation (the logfile):
While running the job status symbol in ADFjobs changes:
The logfile is saved and extended by ADF as it is running.
Right now it is already showing in the ADFtail window, but you could have used any text editor.
Now wait for the calculation to finish:
Wait until ADFtail shows 'Job ... has finished' as last line In the dialog that pops up, click 'Yes' to update the geometry
Now close ADFtail by using the File → Quit menu command:
In the window showing the logfile (the ADFtail window): Select the File → Quit command
In the ADFjobs window, note that the job status icon has changed to indicate that the job is ready:
ADF has created a couple of data files and a couple of text files. You can check this in the ADFjobs window:
Click on the ADFjobs window
Note that the Local files are visible because earlier in the tutorial you already clicked on the triangle opening the details for this job.
The .adf file is the input as saved by ADFinput.
The .run file is the run script, also saved by ADFinput.
To enable specific options of your ADF job, changes to the text of the .run script can be made opening it via double-click. A separate text editor window pops up after the double-click (the non-default text editor program can be defined via the $EDITOR variable passed to ADF-GUI). When you run your job, your edited .run file will be used. Note that when you save your input via ADFinput again, your .run file will be overwritten (and thus you loose any changes you made when editing the .run file).
The .job file is a script file based on the .run file, and is created by ADFjobs. It contains all commands required to run your job, and is actually run by ADFjobs when you use the Run command.
The .pid file is a file that contains your job status and configuration. It is used by ADFjobs.
The other files are produced by ADF. The .err file contains error messages, the .logfile shows the progress of the calculation, the .out file is the main (text) output file, and the other files are binary files that store results of the calculation.
ADF has optimized the geometry, and we can use the ADFmovie module to visualize the progress of the optimization. So let's start ADFmovie using the SCM menu in your ADFinput window:
Select the SCM → Movie command in ADFinput
The ADFmovie module will display a movie of the geometry optimization.
To step through the different geometries, use the slider below the picture. Clicking in the slider will single step through the frames. Alternatively you can grab the slider and move to the frame you wish, or you can use the left and right arrow keys to single step through the frames.
You control playing with the buttons. When your mouse pointer is above any of the buttons, and not moving, a balloon will pop up showing what that particular button will do.
It is often nice to have a graph of the energy as function of the geometry step. ADFmovie can show such graphs easily:
Use the Graph → Energy menu command
You can even show several graphs for different properties at the same time:
Use the Graph → Add Graph menu command Select two carbon atoms by shift-clicking on them Use the Graph → Distance, Angle, Dihedral menu command
Now you have two graphs. One of them is the 'active' graph. When you make a new graph it will always be the active graph. You can also make a graph active by clicking on it.
When you select a property from the Graph menu (Energy, Distance and so on) that property will be plotted in the active graph.
You can also multiple curves in one graph, if possible: one property per Y axes. You may have several curves on the same Y axes if they are using the same unit (all Angstroms for example):
Select two carbon atoms and one oxygen atom by shift-clicking on them Use the Graph → Distance, Angle, Dihedral menu command
Another feature is that you can click on a point in one of the graphs. It will be marked, the movie will jump to that particular step, and if you have more then one graph the corresponding point(s) will also be marked in the other graphs.
To rotate, translate, or zoom the picture, use your mouse, just as in ADFinput.
Use the slider to go to a frame in the middle of the optimization
Selecting atoms provides information about atoms, bonds, etc. The information will be updated when you go to another point in the movie (a different geometry). You can see examples of these in the pictures above.
You can also show this information in the 3D window:
Use the View → Geometric Info → Angle menu command
The angle will be visually added to your molecule :
In the ADFmovie window: select File → Quit
Select your job in the ADFjobs window by clicking on the job name Select the SCM → Levels command
ADFlevels will start and show a diagram of the energy levels of the ethanol molecule.
In the diagram you can see from what fragment types the molecular levels are composed.
Move the mouse around, above different levels, without clicking
Balloons will pop up with information about the level at the mouse position.
Click and hold on the HOMO level of the molecule
In the pop-up that appears, select 'Show Labels' Click and hold on the HOMO of the O fragment type In the pop-up, select 'Show Labels'
To actually see the orbital, select the orbital from the top of the pop-up menu:
Click and hold on the HOMO level of ethanol Select the '10A' command
A window with a picture of the orbital should appear.
You can move (rotate, translate and zoom) the orbital with your mouse.
Close the window showing the orbital: File → Quit (in the window displaying the orbital)
Select SCM → View
ADFview will start up and show a picture of your molecule:
You can use ADFview to visualize all kinds of 'field' related properties: densities, orbitals, potentials, etc. You actually have already used it before: the picture of the orbital that was created using ADFlevels was shown by ADFview.
Use the mouse to rotate, translate or zoom, as in ADFinput.
Select the Add → Isosurface: Colored command
Below the picture a control line will be created. ADFview creates one such line for all visual items and special fields (surfaces, cut planes, calculated fields, etc.) that you add.
In this particular case the control line contains two pull-down menus that you use to select the fields that you want to visualize.
From the first pull-down menu in the control line, select Density → SCF
From the second pull-down menu in the control line, select Potential → SCF
To demonstrate some other possibilities of ADFview, do the following:
Select the Properties → HOMO command Click on the check box in the FIRST control line to hide the density Rotate the molecule to get a good view
Select the Add → Cut plane: Colored command In the new control line, press on the pull-down menu and select Density → SCF Click the check box in front of the 'Isosurface: Double' line to hide the HOMO Select the Carbon and Oxygen atoms (three atoms) Click the 'with atoms' button next to the 'Position plane' option (thus the button, not the check box) Click the check box in front of the 'Isosurface: Double' line to show the HOMO Select the Fields → Grid → Medium command Click Yes to recalculate the fields Rotate your molecule to get a good view
You can save the picture you create using the Save Picture menu command:
Select File → Save Picture ... Enter the name (without extension) of the file you want to create Click Save
A picture with the (file)name you specified has been created.
You might want to explore some more of the possibilities of ADFview on your own. Many different properties can be visualized as you probably have noticed in the pull-down menus. Be careful with activating Anti-Alias: it makes the pictures even better, but also slows down ADFview very very much.
The output file (tutorial.out) is a plain text file. You can view it with your favorite text editor (or text viewer). You can also use the ADFoutput GUI module which provides a convenient way to check the results:
Select the SCM → Output command Select the Iterations → Geometry Cycles command
The ADFoutput program will start showing the results of your calculation, and via the menu you jumped to the first section with geometry details:
You can use the menus to go to different parts of the output file, or you can just use the scroll bar. If a menu option is shaded, this means that no corresponding section of the output is available.
Clicking an output section title highlighted in blue will skip to the next section with the same title, if present.
As we are now done with tutorial 1, close all windows that belong to this tutorial:
Select the SCM → Close All command in any ADF-GUI window
All open windows from the ADF-GUI will be closed.
In the first tutorial you have learned how to construct a molecule by building it out of atoms. That may be a complex task for bigger molecules. ADFinput has other ways to build molecules.
The quickest is to search for a molecule inside ADFinput, and use it if it is available.
Another way is to search for the molecule on the Internet, and use either the xyz coordinates or the SMILES string
Or you can build it using the structure tool in ADFinput. As a more realistic example using the structure tool, you will build a small peptide chain. Then you will learn how to use the predefined metal complex structures. You will learn how to set up your own structures library.
Finally, you can combine the crystal tools to cut molecular systems out of crystals. As an example we will make a sphere of Cu atoms.
For this tutorial we again prefer to work in the Tutorial directory:
cd $HOME Start ADFjobs Click on the Tutorial folder icon Start adfinput via SCM → ADFinput
The quickest solution: search (and find) ethanol.
Press ctrl/cmd-F to activate the search box (or click the search icon in the panel bar) Enter 'ethanol' as search text (without quotes) Click on the 'Ethanol (ADF)' match Rotate to get a good view
Your ethanol is ready. The (ADF) in the search results mean that the molecule has already been optimized by ADF, using the BP86 XC potential with a TZP basis set and small core.
For the next demonstration we need a file with the xyz coordinates of ethanol. You can make such a file using the Export Coordinates option:
Use the File → Export Coordinates... menu command Enter 'example.xyz' as filename Click Save
To import a molecule if you have its structure as xyz file (with element types), you can either use the File → Import Coordinates... or the Edit → Paste command.
Use the File → New menu command in ADFinput Click 'No' when asked if you want to save your changes Click in the ADFjobs window to activate it Use the Job → Refresh List menu command (or press F5) Click the triangle in front of example to show the example.xyz file ctrl-double click on the .xyz file (listed in Local files)
This will open a text editor showing the contents of the the example.xyz file. The editor used depends on your operating system, but normally you will be able to select and copy text.
Select all text in the ethanol.xyz file Copy it Click in the ADFinput window to activate it Paste the xyz coordinates (ctrl/cmd-V or Edit → Paste)
You should again get the ethanol molecule, just as you have saved it.
ADFinput can also interpret SMILES strings (via OpenBabel). As a demonstration, lets try again with Ethanol:
Use the File → New menu command in ADFinput Click 'No' when asked if you want to save your changes Open a web browser Search for ethanol on wikipedia.org At the right side of the page, click the 'Show' link to show the SMILES Copy the SMILES string ( CCO ) Click in the ADFinput window to activate it Paste the SMILES string (ctrl/cmd-V or Edit → Paste) Click in empty space in the drawing area to clear the selection
Again we have an ethanol molecule. SMILES strings do not contain the 3D structure, it was generated by OpenBabel and is NOT an ADF optimized structure. So normally the next step would be to pre-optimize with UFF (via the cog wheel), and to optimize the geometry with ADF.
SMILES strings work often, but there are also many examples where the resulting structure does not make sense.
As a demonstration on how to use the structure tool, we start by building a methane molecule:
Use the File → New menu command in ADFinput Click 'No' when asked if you want to save your changes Select the C-tool Click somewhere in the drawing area to make a carbon atom Select Atoms → Add Hydrogen , or faster: press the shortcut (ctrl-H or cmd-H)
If your cmd-H shortcut does not work (on a Mac), probably you need to fix your X11 prefereces.
The next step is to add a methyl group, using the structures tool:
Select the Structures tool → Alkyl Chains → Methyl structure (the structures tool is the benzene-like logo)
Notice that the button of the structures menu is glowing, which means that the structure-tool is in use.
Double-click on one of the hydrogen atoms Zoom out if needed (with right mouse button or mouse wheel)
You will see that the hydrogen is replaced by a methyl group.
Note that:
To create ethanol, we need to add a hydroxyl group:
Select the Structures tool → Ligands → OH structure Double-click on one of the hydrogen atoms
Again, the hydrogen is replaced by the structure. In this case, the oxygen replaces the double-clicked atom. The hydrogen is precisely aligned along the C-O bond and points away from the rest of the molecule. This shows you the very general way in which the structures will align according to the bonds in the original molecule and those in the structure. In this case, the hydroxyl group is not immediately orientated as it normally would be in an ethanol molecule:
Pre-optimize: click on the cog wheel
And again we have constructed an ethanol molecule.
ADFinput comes with a many predefined structures. Among them are some typical solvent molecules, so that you can easily add solvent molecules around your system. One of these 'Solvent' structures is Ethanol. Now add this molecule in empty space:
Select the Structures tool → Solvents → Ethanol structure Left-click in empty space near the hydroxyl group
Note that the oxygen is selected. Again, this oxygen is defined through having xyz-coordinates (0,0,0). Next we select the new molecule and orient it with the mouse to a reasonable position:
Use the Select → Select Molecule menu command (or ctrl/cmd-M) Use the mouse to rotate and translate the ethanol molecule to your favorite orientation
This way you can easily add explicit solvent molecules.
Now we will build a small peptide chain as another example using the structures tool.
Select File → New Click 'No' as we do not want to save the setup Select the Structures tool → Amino Acids → AA Backbone structure Place it in the drawing area
There appears a subunit (or actually two) of a basic peptide chain. Notice that one of the atoms is selected, namely the terminal nitrogen. This atom is, again, the 'replacing' atom. In order to extend the peptide backbone, you now have to choose the right atom to be replaced. The obvious choice is the (non double-bonded) terminal oxygen.
Click in empty space to deselect the nitrogen
Press the space bar, which activates the last used option of the 'Structure' tool ('AA backbone' here).
Double click on the terminal oxygen.
You may want to use View → Reset View.
In a similar fashion, you can now replace the hydrogens on the backbone by amino acid side groups of your choice. These can be found in the Structures tool → Amino Acid → AA Side Groups sub-menu.
In the sub-menu 'Metal Complexes' you can find a set of predefined complexes corresponding to commonly encountered geometries. Furthermore, there are a number of ligands to be found, which can be easily used with these metal complexes.
Select the File → New command Click No (do not save changes) Select Structures tool → Metal Complexes → ML6 Octahedral tool and place it in the drawing area
Notice that six dummy ("Xx") atoms have been placed around the metal center in an octahedral fashion.
Select one of the dummy atoms by clicking on it Select Select → Select Atoms Of Same Type menu command
The Ligands structure sub-menu contains a number of ligands which can be used to replace the dummy atoms. The Structure menu can, however, also be reached via the Atoms menu.
Select the Atoms → Replace By Structure → Ligands → CN command Click in empty space to clear the selection Reset the View if needed
Notice that all dummy atoms in the selection are replaced by CN ligands.
In order to use the bidentate ligands, we must start with a bare metal center.
Select the File → New command Click No (do not save changes) Place an iron atom in the drawing area (click the X button in the toolbar to get a menu with all elements) Select the Structures tool → Ligands → Bidentates → Ethylenediamine structure Double-click on the metal atom
You can see that, in this case, the metal atom is not replaced by an atom of the structure, contrary to previous experience, but that the bidentate ligand is simply attached to the central metal atom.
This works because the 'replacing' atom in all bidentate structures is a dummy atom, which has the property that it won't replace an existing atom. The metal atom will simply take over the bonds that existed on the dummy atom in the structure. You can easily verify this when you would place the structure in empty space. Other multidentate ligands are defined in a similar fashion.
Press space bar and double-click on the metal atom
Notice that the second ligand appears opposite the existing one.
To change the relative orientation of two bidentate ligands, we can change the plane angle. The planes are defined by two sets of three atoms, the central one being present in both sets. In this case this will, of course, be the metal atom.
Select, in order, the two nitrogens on ligand one, the metal atom, and the nitrogens on the second ligand. Change the plane angle to 90 degrees using the slider
In this way, you can easily change the environment around the metal from square planer to tetrahedral. This feature works as long as you choose the atoms in right order, and if the defined planes can freely rotate relative to each other.
You can make your own structure library very easily.
By default, user defined structures will be stored in the .scm_gui/Structures directory.
To be able to actually use the structures as described earlier, it is necessary to define one of the atoms as having xyz-coordinates (0,0,0). This will then be the atom that will actually appear at the spot of the atom that is replaced by the structure. If you use the Save As Structure command this will be done for you.
Select the File → New command Click No (do not save changes) Build methane Replace three of the hydrogens by chloride atoms and pre-optimize Delete the remaining hydrogen Select the central carbon atom Use the Structures tool → Save As Structure ... command Enter a name like trichloromethyl Note that the selected atom (currently the C atom) will be used as anchor
The new structure will appear in the structures menu and can be directly used.
Dummy ("Xx") atoms are treated a little different when used in structures. A dummy atom will not replace an existing atom when it is defined as the 'replacing atom'. Instead, the double-clicked atom will remain and will accept the bonds that the dummy atom had in the structure.
Build a methane molecule Replace the carbon atom by a nitrogen atom Select one of the hydrogens and replace it by a dummy atom (the Xx atom type, in the periodic system) Select the dummy atom Save the structure using the Structures tool → Save Structure As... command
Now you can select the structure from the structures menu and directly use it.
Select your new structure from the structures menu Double-click on one of the hydrogens
Notice that the hydrogen atom is not removed and that the NH3 group is attached to it. Similar behavior has been demonstrated with the bidentate ligands, where the dummy atoms are also used.
If you want to clean up your structures, you can use the Structures tool → Manage Structures... command. If you use it, ADFjobs will open and show the contents of your Structures directory. As the structures are just (simplified) .adf files, you can open them using ADFinput. And using ADFjobs you can rename them or delete them.
We start making a Cu crystal, using a super cell so we have many real Cu atoms.
To build the crystal, we need to use the periodic tools. These will work only for programs supporting periodicity, like DFTB.
Start ADFinput (or use File → New in the currently open ADFinput window) Switch to DFTB mode (panel bar ADF → DFTB) Select 'Periodicity: Bulk' Edit → Crystal → Cubic → fcc Click OK in the pop-up-window Edit → Crystal → Generate Super Cell... Click OK in the pop-up-window Uncheck View → Periodic → Repeat Unit Cells, so only one unit cell will be shown
Now we have a block of Cu, with explicit Cu atoms (that is using a super cell). Next we will center this block, and select a sphere of atoms around the origin.
Make sure the origin is in the center of the block: Edit → Set Origin Select → Select Atom Close To Origin Select → Select Within Radius Click OK
Select → Invert Selection Press the Backspace key to delete the selected atoms If it does not respond: click once in the drawing area to focus on it, and press the Backspace key again Switch to ADF: panel bar DFTB → ADF Rotate a little
As you can see, you have a (very small) sphere consisting of Cu atoms in the molecular ADF program:
Obviously, by making a bigger super-cell and selecting atoms within a larger radius you can make bigger spheres.
In this tutorial we demonstrate the working of several tools that combined give you great flexibility to build periodic structures.
If you are not interested in periodic structures (for example, if you only use ADF), you can skip this tutorial.
The instruments at your disposal are:
To use these tools you need to use ADFinput in periodic mode. For example, by switching to the BAND mode, or by switching to DFTB or Mopac and next changing the periodicity.
If you are lucky your crystal structure is in the database. Of course there are infinitely many possible crystal structures, so the database has to be incomplete. Nevertheless, the most common structures are there. NaCl is one of them.
Start BANDinput (or DFTBinput, and switch the periodicity to 'Bulk') Click on the snowflake like pictogram on the toolbar. Select a "Cubic" lattice and then NaCl
Next a dialog pops up where you can change the parameters of the structure, such as lattice constants
In this case there is no need to change anything.
Click OK
More often a crystal is not directly in the list. An example is LiF. It has the same crystal structure as NaCl, but other elements and a different lattice constant, namely 4.01:
Open again the NaCl dialog Change the lattice constant and the elements as shown Click OK
The NaCl crystal structure will be replaced by the new LiF crystal structure.
Some crystal structures have more parameters.
Select "Tetragonal" and then "rutile" from the snowflake tool Click OK
As you can see, it has two lattice parameters that need to be specified. But now also the positions of the oxygens have a parameter "x"
The structure database is very convenient but by necessity incomplete. A more powerful approach is to build 3D crystals from the space group information. Here we give an example on how to build TiO2, also known as rutile.
Open the crystal builder: Click on the snowflake tool and select "From Space Group"
And the following window pops up
Rutile has the symmetry of space group 136
Enter 136 in the "Number" field, and a return
Note how the Browser reflects the change and also how the "Name" and "Lattice" values change
Now set the two lattice parameters as below
We still need to define the atomic coordinates. For starters click on the plus below "Coordinates"
In a book on crystal structures you can find that rutile has two sites occupied. The Ti atom is on the "a" site
Select the Ti atom and select the "a" site
The oxygens occupy the "f" site.
Click on the plus to add a site Change the atom type to "O" and the site to "f"
As you can see in the "Coords" column and the "Parameters" column, this site has an undetermined parameter "x". (It represents a symmetry line for this space group.) In the book you can find that for TiO2 "x=0.3".
Set "x" to 0.3
The final dialog looks like
Press "Apply" and if that looks any good press "Close"
The slicer is a very easy, yet powerful tool to make slabs from any crystal structure.
Select fcc from the "Cubic" crystals
The element and lattice constant are already correct for Cu.
Press "OK" to generate the Cu lattice
Let us invoke the slicer tool to cut out the slab.
The slice tool icon looks like this: 
Click on the slicer tool icon in the toolbar
The following dialog appears
Set the Miller indices to (1,1,1) Select Cartesian Set the number of layers to 3
The "Cartesian" option is needed because the Miller indices are usually thought in the conventional unit cell rather than the primitive (minimal) unit cell.
Press OK
After pressing OK you will see (from the top)
You may want to rotate it or press the side view buttons in the toolbar to convince yourself that it is a three layer slab.
Thus far, we have always created primitive unit cells: the smallest possible. Quite often you want to use a larger, less symmetric unit cell. You can do this for chains, slabs, and crystals, but we will demonstrate how it works for a slab.
We will continue to work with the 3-layer Cu slab:
Select the Edit → Crystal → Generate Super Cell... command
Thus invoking the Super Cell Tool
Here you see how new lattice vectors are expressed in terms of old ones. Because we have a slab this is a 2x2 matrix.
Select the "sqrt(3) x sqrt(3)" option from the "Preset" menu
The matrix has now changed into 2,-1,-1,2.
Press OK
and you get a unit cell with three atoms per layer:
Close BANDinput
First you will construct an ethene molecule and optimize its geometry. As this is basically the same procedure as in Tutorial 1, only short instructions will be given.
Next you will set up the calculation of excitation energies, and let ADF perform the calculation.
Finally, using ADFlevels, ADFspectra and ADFview, you will examine the results.
For this tutorial we again prefer to work in the Tutorial directory.
You know how to start ADFjobs (in your home directory), and move to the Tutorial directory:
cd $HOME Start ADFjobs Click on the Tutorial folder icon
Next start ADFinput using the SCM menu:
Select the SCM → ADF Input command
Note that if you start ADFinput via the SCM menu you will start ADFinput without loading a job. If you wish to open ADFinput for a particular job, click on the 'ADF' button in front of it.
For the View command (and Movie, Levels, and so on) the behavior is different: it will immediately load the job that is selected in ADFjobs.
First we construct an ethene molecule, and pre-optimize its geometry:
Select the C-tool by clicking on the button with the 'C'
Select the Double bond mode Bond tool → Double
(or use the shortcut: press the '2' - key)
Click somewhere in the drawing area to create a carbon atom
Click again to create a second carbon atom (with a double bond)
Click on the just created Carbon atom to stop bonding
Select the Atoms → Add Hydrogen command
Click the cog wheel to pre-optimize
Your ethene molecule should look something like this:
The next step is to optimize the geometry using ADF:
Enter a title in the Title field (like 'Ethene') Select the 'Geometry Optimization' preset
With the proper options selected, now run ADF:
Select the File → Run command Click 'Yes' in the pop-up to save the current input In the file select box, choose a name for your file (for example 'ethene') and click 'Save'
Now ADF will start automatically, and you can follow the calculation using the logfile that is automatically shown.
Wait until the optimization is ready (should take very little time)
The following dialog should appear:
Click 'Yes' to update the coordinates
To set up the calculation of the excitation spectrum:
Adjust the title to something like 'Ethene excitations' Select the 'Single Point' template Use the panel bar Properties → Excitations, CD command to go to the Excitations panel For the 'Type of excitations' option, Select 'Singlet and Triplet'
For the tutorial this set up is fine, normally you would also need to select an XC potential that gives better results, for example SAOP.
Now everything is ready to actually run ADF. Before running we will save the current input in a different file (though this is not really required):
Select File → Save As... Enter a filename (ethene-exci) and click 'Save' Select File → Run Wait for the calculation to finish
The logfile shows you that the calculation has finished, and that indeed the excitation code has been running:
Select SCM → Levels
In this level diagram you can see that the HOMO and LUMO consist mainly of carbon p orbitals. It is also easy to see what orbitals the hydrogens take part in.
Note that the carbon and hydrogen stacks show all carbon and hydrogen atoms at once: they show the fragment type. ADFlevels can also show the individual fragments but when using atomic fragments you will get too many fragments. In this particular case symmetry is used, and since there is only one symmetry unique carbon atom and only one symmetry unique hydrogen atom you still would see only one stack per atom type.
Select SCM → DOS In the ADFdos window: View → Add Graph Select one hydrogen atom
In these plots you can see that the partial DOS for the hydrogen atoms have no contribution to the HOMO. By right clicking on an atom you can also show partial DOS graphs with contribution from selected atoms and selected L-values only.
Select SCM → Spectra
ADFspectra will start and show the calculated excitation spectrum.
When moving the mouse above a peak a pop-up will give you more information about that peak:
Move the mouse above the largest peak and hold it there
The composition of the excitation in terms of orbital transitions is listed in the pop-up. If you move the mouse inside the pop-up and click on a line the corresponding orbitals will be shown using ADFview:
Move the mouse inside the pop-up and click on the line with the major contribution
Close the two windows showing the orbitals using File → Quit in both windows
We will now use ADFview to examine the orbitals. Not just one, but have a look at many of them. To do that ADFview has a simple method to make something similar to an 'orbital selection' panel: you can tear off the orbital menus!
Select SCM → View
Select Properties → HOMO
Click on the field selector pull-down in the control bar for the HOMO (reading SCF_B1.u 1: ...),
select Orbitals (occupied)> → 1 → - - - - - - - - (the dashed line)
The menu will tear off, place it at a convenient location
Now you can just click on the different orbitals in the torn-off menu, and very quickly examine those orbitals:
Click on all of the orbitals in the torn-off menu, one by one, and observe the orbitals Click again on them
When you click the first time on an orbital, its values need to be calculated and then the orbital is shown. When you click for a second time on an orbital in the list, it has already been calculated (indicated by the check mark). And thus it shows immediately.
Close the torn off menu by clicking in its close box
In the same way you can tear off many menus in the GUI, for example to select virtual orbitals. Not only in the ADFview control bars, but also from the main menu bar. Anywhere you see the dashed menu line.
Close ADFview: File → Quit
You can use ADFview to view orbitals etc, but also to have a look at the transition density. At this moment ADFview cannot generate this field automatically, it needs to be calculated first using DENSF. The ADFinput defaults are such that if you perform an excitation calculation, DENSF will automatically calculate the corresponding transitions densities. So we will start ADFview and next load the .t41 file generated by DENSF:
Select SCM → View Select File → New Select File → Open In the file-select pop-up, choose to see all .t41 files using the pop-up menu at the bottom Select the .t41 file from your calculation (ethene-exci.t41) Click 'Open' Select Add → Isosurface: Double (+/-)p
In the field pull-down menu (in the control line for the double isosurface) you will now find an entry 'Transition Density', and if you select that you will find all transition densities that are available.
In this case lets select the transition density that belongs to the biggest peak: the Singlet-Singlet 1B3u excitation:
Select the Transition Density → TransDens_SS_B3.u_Fitdensity_1 field Change the iso value to 0.003 Rotate the molecule a little
The resolution is rather coarse, you can use interpolation to get a better picture. Changing the grid has no effect in this case as you are visualizing a .t41 file, which is just data on a grid (in binary form).
Fields → Interpolated Select the Transition Density → TransDens_SS_B3.u_Fitdensity_1 field in the I-1 line In the Isosurface: Double line, change the field to I-1 (in the Other category in the field selector)
Using the output browser you can find all details about your excitation calculation. Use the menu to jump to the relevant part of output:
Select SCM → Output Select Response Properties → All Singlet-Singlet Excitation Energies
You might get minor differences in the results as the default integration accuracty is only 4.
To close all modules for your excitations calculation at once, use the Close command from the SCM menu:
Select SCM → Close
Close will close all open modules that have your current job loaded, except ADFjobs. The Close All command will close every ADF-GUI module, including ADFjobs:
Select SCM → Close All
All ADF-GUI modules (and BAND-GUI modules if any) should be closed.
For this tutorial we again prefer to work in the Tutorial directory:
cd $HOME Start ADFjobs Click on the Tutorial folder icon Start ADFinput using the SCM menu
Next we create an ethane molecule. This is almost identical to making ethanol:
Select the C-tool Click once to make a carbon atom Click again to make a second carbon atom Select Edit → Add Hydrogen And do it again: ... add hydrogen Click the pre-optimize button (the cog wheel) Click the Symmetrize button (the button with the star, on the bottom toolbar)
You need to add hydrogen twice: the first time one carbon atom was selected. Thus, the 'Add Hydrogen' command added the hydrogen to that selected atom only. By repeating the 'Add Hydrogen' command the hydrogens are added in all possible places.
Alternatively, you can, after placing the second atom, click on that atom to finish using the C-tool. In this case none of the carbon atoms will be selected and you can easily add all the hydrogens at the same time.
The symmetrization should have enforced perfect D3d symmetry, using the Symmol program. The Symmol program has been developed by Pilati and Forni, see the balloon help on that button for details..
Now optimize the geometry using ADF, as in the second tutorial:
Enter a title (like 'Ethane')
Select the 'Geometry Optimization' preset
Select File → Run
Save your input with a nice name ('ethane')
Wait for the calculation to finish (check the logfile)
In the dialog that appears, click Yes to import the optimized geometry
Now we will set up and run the frequencies calculation. It is important to start with the optimized structure, so make sure you did import the optimized coordinates (the last step of the previous instructions).
Select the 'Frequencies' preset
For the tutorial we will use the default XC potential. For better results you should select a better XC potential.
Select File → Run Click 'OK' to save over the previous version Click 'Yes' when ADFjobs warns that results are already present Wait for the calculation to finish
When the calculation is finished the end of your logfile will look something like this:
At the end ADF calculates the force constants for the vibrations.
Select SCM → Spectra
The IR spectrum is shown. You can save a picture to use in other programs using the 'Save Postscript' command from the 'File' menu. Alternatively, you may export the xy coordinates of the spectrum to a plain text file for you own graphing application.
If you move the mouse pointer above the spectrum you will get more information about the peaks in pop-up windows:
Move the mouse pointer above the largest peak and keep it there
The list of normal modes contributing to the IR peak will show up. By clicking the mouse button at this point, all these modes will be animated, each opened in a separate ADFmovie windows. Alternatively, pressing and holding the mouse button down, a list of these modes with their symmetry labels will show up:
Press and hold down the mouse on the largest peak
Choosing one of the modes from the the pop-up list, you can visualize only the corresponding mode.
Select one of the modes in the pop-up menu
ADFmovie should start and show the corresponding normal mode.
You can adjust how the normal mode is visualized. You can make the displacements larger or smaller. The commands are in the Scale Displacements menu located in the View menu, but using keyboard shortcuts is much easier here:
Use ctrl-L (or cmd-L on a Mac) repeatedly to make the displacements larger Use ctrl-K (or cmd-K on a Mac) repeatedly to make the displacements smaller Use the View → Displacement Vectors to visualize the normal mode with vectors
You can make the vectors larger or smaller using the same ctrl/cmd-L/K
Sometimes you wish to examine many different normal modes. Obviously you can do this as just described, but another way is by selecting the mode of interest from the menu. And to make it more convenient to check many of them, you can tear off these menus, just as you did with the orbital selector menu in ADFview.
In ADFspectra: File → Quit, to close all normal modes that might be showing Start ADFspectra again SCM → Spectra Click on the E1u-1 menu, and select the dashed bar (- - - -), thus NOT one of the modes Position the torn off menu in a convenient location
Click on each of the modes in the torn of menu in turn Observe the movie windows showing the normal modes
As you will find out, this is an extremely nice and convenient way to examine many different modes.
We are done with this tutorial, so close all open ADF-GUI windows:
SCM → Close All
This tutorial consists of several steps to study the isomerization reaction of HCN:
cd $HOME Start ADFjobs Click on the Tutorial folder icon Start adfinput using the SCM menu Draw an HCN molecule (first the N, next the C and finally an H atom) Select the C-N bond and make it a triple-bond Pre-optimize the geometry
You should get a linear molecule:
Select the "Geometry Optimization" preset Select the DZP basis set Select File → Run, give it the name HCN_GO
The geometry will be optimized, using a DZP basis set.
Click "Yes" when asked to read new coordinates from the HCN_GO.t21 file Check the C-N and C-H distances
They should be about 115 and 108 pm (1.15 and 1.08 Angstrom), respectively.
Write down the value of the bonding energy printed at the end of the calculation in the ADFtail window
The HCN molecule has an CNH isomer. There is an energy barrier between these two states. We are going to find the transition state and calculate its height.
To find a better starting point for the transition state search we will perform a linear transit calculation as a rough approximation of the reaction path. We will vary the H-C-N angle in steps between 40 and 140 degrees and let ADF optimize bond lengths at each angle.
To set up the linear transit calculation:
Select the 'Linear Transit' preset Click on the "..." button next to the GeometryOptimization task to go to the 'Geometry Constraints and Scan' panel Select all the atoms
You should see '+ N(1) C(2) H(3) (angle)' note in the right panel now:
Click the '+' button to add the angle constraint
Now '- N(1) C(2) H(3)' and the two fields as limits for the degree parameter appear.
Enter '140' and '40' in the 'degrees' fields
ADF will have trouble running the current set up because the HCN molecule is perfectly linear. So we will help ADF by changing the angle to 140 degrees, the same as the first point of the LT scan.
Use the slider to change the HCN angle to 140 degrees
The set up is complete. Now we will run the LT calculation, but we will save it with a new name as we wish to keep the results of the HCN_GO calculation:
Use File → Save As to save the file as 'HCN_LT' Run the calculation
Running might take a few minutes. When the run is finished:
Click "Yes" when asked to read new coordinates
You will see the last geometry, close to CNH. To see how geometry was changing during the LT run, use ADFmovie:
Select the SCM → Movie command Select all atoms (use shift-drag to make a rectangle around the atoms) Use the View → Align Screen command to make sure you can see all atoms Select the Graph → Energy command Select the Graph → Distance, Angle, Dihedral command (to get the angle graph, as three atoms are still selected) Use the View → Converged Geometries Only command Zoom in to get a better view of the molecule Press the Play button (the right-pointing triangle in the controls at the left bottom of the ADFmovie window)
You will see the hydrogen atom moving from C to N. You will also see a graph of the energy as function of the LT steps. As the movie is playing a dot shows the corresponding position in the graph.
Somewhere along the path, there is a transition state we are looking for. Remember that you needed to use the 'Optimized Geometries Only' command to filter out all the intermediate geometry step, so that you get only the converged geometries for each LT step.
In the graph, click (without moving!) on the top of the energy graph Alternatively, use the arrow keys (cursor keys) to move between different steps, or use the slider Check which geometry has the maximum energy
You should find that at around an angle of 60-75 degrees the maximum energy is reached. This is Frame 6 (the 7th LT step):
You can find this information also in the output file:
Select the SCM → Output menu command In the ADFoutput window select the Other Properties → 'LT Path command
You will see that indeed the geometry number 7 (corresponding to Frame 6 in ADFmovie) has the highest energy. In this particular example the choice of the angle is not very important, but in general you will always want to get the best approximation for the transition state available.
We will now prepare the search for the transition state starting from this geometry:
Click in the ADFmovie window Make sure frame 6 is selected Use the File → Update Geometry In Input command
The geometry of HCN in your ADFinput window will be updated to match the geometry currently selected in the ADFmovie window:
In general, it is important to have a good starting Hessian with one imaginary frequency when performing a TS search. We are going to create such a Hessian by doing a quick frequencies calculation:
Select the "Frequencies" preset (from the Main panel) Set the integration accuracy to 4 Select the 'Geometry Constraints and Scan' panel (via the ... button or via the panel bar) Remove the Angle Constraint: click on the '-' in front of the angle constraint Save the molecule with as 'HCN_Freq1' (Save As) Run the calculation
The frequency calculation is now in progress and will run very fast. When it has finished:
Select the SCM → Spectra command
If everything was done correctly, you should see a spectrum with three peaks: two in the range of 2000 - 3000 1/cm and one peak in the range of negative or very low positive values. A negative frequency value actually means that it is an imaginary frequency.
Click with your mouse on the peak corresponding to the imaginary frequency
The ADFmovie program will start displaying the normal mode corresponding to the frequency. Check that the frequency indeed corresponds to the H atom moving parallel to the C-N bond.
The result file, HCN_Freq1.t21, has an initial geometry for our transition state search. It also contains a Hessian matrix (produced with the frequencies calculation) that can be used to kick-start the TS procedure.
Bring ADFinput with the HCN_Freq1 calculation to the foreground Select the 'Transition State Search' preset ctrl/cmd-F and search for 'restart', select the 'Files (Restart)' panel Click the file select button (looks like a folder) in front of the empty 'Restart file:' field Select the HCN_Freq1.t21 file and click 'Open':
Save the set up as HCN_TS (Save As) Run the calculation
After the calculation has finished (again very fast), you will be asked to read the new geometry from the results file HCN_TS.t21:
Answer "Yes" to import the latest geometry Make a note of the bond energy for the transition state (visible in the logfile)
ADFinput will now display the transition state geometry.
If you compare the bond energy with the bond energy of the optimized HCN molecule from the first calculation, the difference should be about 1.9 eV. Also check that the geometry makes sense: the C-H and C-N distances should be around 1.20 and 1.19 Angstrom and the H-C-N angle should be about 70 degrees.
After every transition state search it is good practice to verify that you indeed have one and only one imaginary frequency. For this we will repeat the frequency calculation at the TS geometry:
Make sure you have HCN_TS open in ADFinput Select the "Frequencies" preset (from the 'Main' panel) Save with name HCN_Freq2 Run
The calculation is running, should not take much time. After the calculation has finished:
Select the SCM → Spectra command
You will be presented with an IR spectrum of the molecule featuring three bands roughly located at 2550, 2050, and (imaginary) -1040 1/cm.
Click on the band at -1040
You should get a movie showing the normal mode corresponding to this frequency, telling us that this is indeed the reaction coordinate we are studying.
ADF can follow the minimum-energy path from the transition state to one or the other product. The method used in ADF for this is called Intrinsic Reaction Coordinate (IRC). You may want to skip this part as the calculation might take some time to complete.
Bring HCN_Freq2 in ADFinput to the front Select the "IRC" preset Go to the 'Intrinsic Reaction Coordinate (IRC)' panel (in Model, or via the search, or by clicking on the '...' button)
This panel allows you to specify various parameters for the IRC method. The most important parameter is the direction to follow. The choice is more or less arbitrary. By choosing "Forward path" or "Backward path" will lead you to one or the other product but it's hard to tell which of the two in what case. We will calculate both paths at once, so we do not need to change the default.
Save as HCN_IRC Run
After some minutes the calculation will finish. You can use ADFmovie to view the IRC path. Of course you need again to make sure to show the converged geometries only:
Select the SCM → Movie command Select the View → Converged Geometries Only command Select the Graph → Energy command Select all atoms Select the Graph → Distance, Angle, Dihedral command
From this movie you can see the IRC path, and the energies at the most interesting points. As we have calculated the forward and backward path in one run, in the movie you will first see the IRC in one direction (forward), and next resuming in the other direction (backward).
You can also examine some properties along the IRC path by studying the output file:
Select SCM → Output (this might take a while, be patient) Use the Other Properties → IRC Path menu command
You will see a table with the properties along the forward path. To get the backwards path:
Click on the blue header 'Dist from TS ...'
The output browser should jump to the next section with that header, which is the table for the backward IRC path.
Finally, at the end of this tutorial you will have many open windows. To close all ADF-GUI related windows at once, you may use the SCM → Close All command.
This tutorial consists of several steps:
First create a TlH (thallium hydride) molecule with a bond length of 1.87 Angstrom (the experimental bond length):
Open ADFinput and draw a TlH molecule. Select the Tl and H atoms Use the slider to set the distance between the atoms 1.87 Angstrom
Next we will set up the calculation. The following details need to be set:
Clear the selection (click in empty drawing space) Set the title to 'TlH spin-orbit coupled' Select the BP GGA as XC potential in SCF Select the 'Spin-Orbit' relativistic option Select the basis set 'TZ2P' Select the core type 'None'
The Main panel will now look like:
We are going to perform a fragment analysis as a trick to get a diagram that makes it very easy to compare scalar and spin-orbit relativistic results.
Fragment calculations are based on regions, which are just collections of atoms. So we start by making a region:
Select both atoms Use the panel bar Model → Regions command Click the '+' button to add a new region Change the name of the new region (Region 1) to TlH_SR
You have now defined a region containing all atoms, with name TlH_SR.
Use the panel bar Multilevel → Fragments menu command Click the 'Use fragments' check box
Use the File → Save menu command Enter the name 'TlH_SO' in the 'Filename' field Click 'Save'
Now you have saved your current options and molecule information.
As we have set up a fragment calculation, also the .adf and .run files for the fragment have been saved. Lets study what options are used for the fragment in ADFinput:
Make sure the 'Fragments' panel is still the current panel Click on the 'Open' button (the dot) for the TlH_SR fragment
A new ADFinput window will also appear with the name 'ADFinput: TlH_SO.TlH_SR.adf'. This is the name of the molecule, a dot, and the name of the fragment. The fragment should have the 'Scalar' relativistic option selected, as that is required when the results will be used as a fragment. The other options are identical to what you set for the main molecule.
Now close this ADFinput window:
Select the ADFinput window with the name 'ADFinput: TlH_SO.TlH_SR.adf' Select File → Quit
We are now ready to run the calculation:
Select the ADFinput window with the name 'TlH_SO.adf. Select File → Run
Now two calculations will run: first the building fragment (using the scalar relativistic option), and next the version including spin-orbit coupling. You will see the two logfiles. Wait until both calculations have finished:
Wait until ADFtail shows 'Job ... has finished' as last line Select File → Quit Repeat for the second ADFtail, thus closing both logfiles
To see the effect of the spin-orbit coupling we will first look at the energy level diagram:
Select the ADFinput window with the name 'TlH_SO.adf. SCM → Levels Select View → Labels → Show Press and hold the Right mouse button on the stack name 'TlH_SO', Click on 'Zoom HOMO-9 .. LUMO+9'. Next try to zoom using a drag with the right mouse button, or using the scroll wheel. Do this such that only levels between -0.1 and -0.7 eV are shown. You can move the levels vertically by dragging with the left mouse button.
You can see that the spin-orbit coupling is important to split energy levels.
Especially for the Tl core levels the spin-orbit coupling is more important than the ligand field splitting. Compare the 8pi, 13sigma, 4delta orbitals (close to 5d atomic Tl orbitals) with the 11j3/2, 20j1/2 spinors (close to 5d3/2 atomic Tl spinors) and 5j5/2, 12j3/2, and 21j1/2 spinors (close to 5d5/2 atomic Tl spinors).
If you press and hold the right mouse button on one of the levels, you can select a spinor. That spinor will be shown. You can also show all spinors (in the case of a degenerate level) at once.
The energy diagram of the scalar relativistic fragment calculation shows the atomic contributions to the scalar relativistic levels.
Bring ADFjobs to the front Select the TlH_SO.TlH_SR job (the scalar relativistic fragment) Use the SCM → levels command Select View → Labels → Show Press and hold the Right mouse button on the stack name 'TlH_SO.TlH_SR', Click on 'Zoom HOMO-4 .. LUMO+4'. Next zoom and move the levels using a right mouse drag and or scroll wheel.
Visualization of spinors is conceptually more difficult than visualization of orbitals.
A spinor Ψ is a two-component complex wave function, which can be described with four real functions φ: real part α φαR, imaginary part α φαI, real part β φβR, imaginary part β φβI:
| Ψ = ( |
φαR + i φαI φβR + i φβI |
) |
The density ρ is:
ρ = Ψ† Ψ
The spin magnetization density m is:
m = Ψ† σ Ψ
where σ is the vector of the Pauli spin matrices. A spinor is fully determined by the spin magnetization density and a phase factor eiθ, which both are functions of spatial coordinates.
In ADFview one can visualize the (square root of the) density and spin magnetization density, however, the phase factor eiθ is summarized only with a plus or minus sign.
For this tutorial we have a small molecule, and a fine grid is chosen for better visualization.
Select the ADFinput window with the name 'TlH_SO.adf. Select SCM → View Rotate the molecule, such that one can see both atoms. Select Fields → Grid → Fine Select Add → Spinor: Spin Magnetization Density In the new control line at the bottom, use the field select pull-down menu and Select Orbitals (occupied) → 3 → SCF_J1/2:1 22: 1.0 ....
The arrows in this picture are in the direction of the spin magnetization density m. All arrows are approximately in the same direction, which means that this spinor is an eigenfunction of spin in this direction of the arrows. In fact this 22 j1/2 spinor is almost a pure α orbital. The arrows are drawn staring from points in space where the square root of square root of the density is 0.03. The color of the arrows are default red or blue, indicating minus or plus for the phase factor eiθ.
The (square root of the) density and the approximate phase vector eiθ can also be viewed separately:
Select Add → 'Isosurface: Double (+/-) Select Orbitals (occupied) → 3 → 'SCF_J1/2:1 18: 1.0 .. Hide the spinor (uncheck the check box at the left of the Spinor label)
This spinor 18j1/2 is almost a pure 5p1/2 Tl spinor. A p1/2 atomic orbital has a spherical atomic density, but a spin magnetization density which is not the same in each point in space.
In the control line with 'Spinor', press on the pull-down menu and Select Orbitals (occupied) → 3 → SCF_J1/2:1 18: 1.0 ... Show the 'Spinor' (check the left check box for the spinor line) Hide the 'Double Isosurface' (uncheck the left check box for the double iso line) Hide the atoms: View → Molecule → Sticks
The calculation of the atomization energy is not a simple problem in DFT. Spin-orbit coupling is an extra complication. In this paragraph a way is presented how to calculate the atomization energy using spin-polarized calculations in the non-collinear approximation.
If you wish, you can skip the rest of this tutorial.
To calculate an atomization energy we need to calculate the atoms also including spin-orbit coupling. The easiest way is to start with the TlH_SO.adf file and change this to an atomic file.
Since the Tl atom is an open shell atom for an (accurate) atomization energy we need to do an unrestricted calculation. The best theoretical method is the non-collinear method. Note that the 'Spin polarization' field is not used in the non-collinear method.
Select the ADFinput window with the name 'TlH_SO.adf Delete the H atom (select it and press the backspace key) Use the panel barMultilevel → Fragments command Uncheck the 'Use fragments' option Use the panel barModel → Regions command Remove the TlH_SR region (click on the - button in front of it) Select 'Main' panel Enter the name 'Tl spin-orbit coupled' in the 'Title' Menu' Check the 'Unrestricted:' box Select the Relativity panel (search for relativity) Select 'NonCollinear' from the 'Unrestricted Options' Menu Select File → Save As Enter the name 'Tl_SO' in the 'FileName' field Click on 'Save' Click OK to acknowledge the warning about fractional occupation numbers
Now we want to actually perform the calculation for the Tl atom
Run the calculation: File → Run Wait until ADFtail shows 'Job ... has finished' as last line In the window showing the logfile (the ADFtail window Tl_SO.logfile): Select File → Quit
Basically we can follow the same steps as for the Tl atom, but in this case we will start with Tl_SO.adf file and change this.
Select the ADFinput window with the name 'Tl_SO.adf Select the 'Main' panel Enter the name 'H spin-orbit coupled' in the 'Title' field Select the Tl atom Use the Atoms → Change Atom Type → H Select File → Save As... Enter the name 'H_SO' in the 'Filename' field Click on 'Save' Select File → Run Wait until ADFtail shows 'Job ... has finished' as last line In the window showing the logfile (the ADFtail window H_SO.logfile): Select File → Quit
The atomization energy including spin-orbit coupling is a combination of several terms.
Select the ADFinput window with the name 'TlH_SO.adf. Select SCM → Logfile Write down the value of the bonding energy printed at the end of the calculation in the ADFtail window. (should be around -1038.62 eV) Select File → Open Select the file 'TlH_SO.TlH_SR.logfile' Write down the value of the bonding energy printed at the end of the calculation in the ADFtail window. (should be around -3.84 eV) Select File → Open Select the file 'Tl_SO.logfile' Write down the value of the bonding energy printed at the end of the calculation in the ADFtail window. (should be around -1039.32 eV) Select File → Open Select the file 'H_SO.logfile' Write down the value of the bonding energy printed at the end of the calculation in the ADFtail window. (should be around -0.95 eV)
The atomization energy including spin-orbit coupling is in this case, the bond energy printed in the TlH_SO.logfile plus the the bond energy printed in the TlH_SO.TlH_SR.logfile minus the bond energy printed in the Tl_SO.logfile minus the the bond energy printed in the H_SO.logfile. (approximately -1038.62 - 3.84 + 1039.32 + 0.95 = -2.19 eV, experimental number is close to -2.06 eV.)
In this tutorial the basic concepts of setting up multi-level calculations in the ADF-GUI will be demonstrated. In most cases one would use a multi-level method for big systems: handle the full system with a fast method, and use ADF to study a particular region of interest with more detail. As big systems will take too much time for a tutorial, the concepts will be shown with very small toy systems that are not typical applications.
For all multi-level calculations you will need to define regions.
Within ADFinput, a region is a collection of atoms. You manage your regions with the 'Regions' panel in ADFinput. It allows you to define new regions, modify existing regions, and to apply some commands on particular regions.
As a first example, we will work with an acetone molecule (CH3COCH3) to demonstrate how to use QUILD.
Start ADFinput, and build an Acetone molecule (or search for it ...) Select one of the methyl groups Use the panel bar Model → Regions command Click on the '+' button to add a new region Change the name 'Region_1' into 'Methyl' (the name is editable)
You have just defined a new region, with a name 'Methyl'. In your molecule window you can see what atoms are part of this region: they are enclosed in slightly ghost-like spheres. Note that there are now (at least) two regions define: the region that you defined, and a region called 'All' that is always present. Obviously, the All region includes all atoms. If you created the acetone molecule via the search command, you may have a third region that was automatically made.
Now we will define yet another region: the rest of the molecule. One way to do this is just to select the atoms that should be part of it, and pressing the '+' button as you did before. However, to demonstrate some of the things you can do with regions we will do it in another way:
Make sure you have only the All and the Methyl region; if you have more, press the '-' button in front of the region Click the check button in the 'All' region Click the '+' button to create a new region, as you did before Click the check button in the 'Methyl' region Click the '-' button in the 'Region_2' line, on the right-hand side
Basically, what you just did: select all atoms, make a new region containing all atoms, select the atoms in the Methyl region, remove the selected atoms from the new region. By clicking on the select buttons in the regions you can easily verify that your regions are now as they should be.
You can easily change what regions, and atoms within that region look like:
Click in empty (drawing) space to clear the selection Click in the check box in front of the 'Methyl' region name to uncheck it
You should observe that the ghost-like region visualization disappears. Please turn it back on:
Click in the check box in front of the 'Methyl' region name to check it Click on the right arrow at the end of the 'Region_2' line, and use the 'Sticks' command from the menu that appears
Now anything in 'Region_2' will be visualized as sticks only. Obviously you could also select any of the other display options.
Once you have defined your regions, it is easy to set up the QUILD calculation:
Use the panel bar ADF → Quild menu to select the Quild panel Click the '+' button to add an action Click the '+' button to add a second action In the first action, select DFTB in the 'Use' field In the second action, select ADF in the 'Use' field
The first action ('Total') defines what to do with the full molecule. It normally will apply to the whole molecule, and thus the 'All' region is preselected. In the 'Use' field you have specified how to treat the whole molecule: with DFTB.
The second action ('Replace') tells QUILD to replace the DFTB result for the selected region with results from another method. The region for which we want to do this is the 'Methyl' region, and it happens to be automatically selected. You can use the region pull-down menu will to select another region, and it offers you a short-cut to make a new region.
In the 'Use' field of the second action you have selected what to use as a replacement method: ADF.
In the 'Remove' field it should be specified that you wish to remove the DFTB results for this region. ADFinput will enter this automatically when you save your job. You can also set it manually if you wish.
The Quild panel offers you to set some details for the QUILD calculation. Normally the defaults should work fine.
Save your set up
Below the Quild panel you will now find three tabs: 'Quild', 'DFTB 1', 'ADF 1' and 'DFTB 2'. These tabs allow you to set up the different parts of the calculation. Right now you could make adjustments of the global QUILD settings. If you press on the 'ADF 1' tab, you will have the option to set the details of the ADF calculation (for the 'Methyl' region). And if you click on the 'DFTB 1' tab you can set up details of the DFTB part of the calculation. The 'DFTB 2' tab is the DFTB calculation on the 'Methyl' group that will be removed from the full system. Normally the set up for this calculation is identical to the full system (DFTB 1), but in some special cases you will need to modify it. See the QUILD manual for details.
Click on the 'ADF 1' tab Look through the different panels, to see what options ADF will use. Do NOT make changes! Click on the 'DFTB 1' tab Click on the 'DFTB 2' tab Click on the 'Quild' tab
Now let's run this calculation:
File → Run Click 'No' when asked to update the geometry
When your calculation is finished you can view the resulting optimization using ADFmovie:
Use the SCM → Movie command Add a graph: Graph → Energy Add a second graph: Graph → Add Graph Show the energy of the ADF-subsystem on the second graph: Graph → Quild Energies → job 2 : ADF ...
You can also open the output file using the SCM → Output command. The other visualization tools can not be applied to the full QUILD results, but they can be used to examine the result of the ADF calculation (on the Methyl region). This is done in quildjob.2:
Show the ADFjobs window Show the QUILD job details (click on the triangle to show the details)
Using the View menu command you will try to open the .t21 result file for the QUILD job. That will not work, we need to view only the .t21 file of the adf sub-system. To do this, we first open this result file in the KFBrowser. This tool allows you to inspect details of the binary KF files. Next, use the View command in the SCM menu of the KFBrowser to open that specific file in ADFview:
Double-click on the .results/quildjob.2.t21 file In the KFBrowser window, SCM → View In the ADFview window Remove and Guess bonds (this is a work-around a bug that messes up the bond display) Visualize the HOMO in ADFview Click on the Isosurface:Double pull-down menu and use the Show Details command Change the Opacity to 70
Note that QUILD has added a dummy hydrogen to cap the broken bond.
In the ADFview window, use the SCM → Levels command
Now you have the ADF result file for the methyl group visible in ADFview and ADFlevels. With ADFview and ADFlevels you can examine the ADF results as usual.
SCM → Close All
To demonstrate how to set up a QUILD calculation using ADFinput, we will use ethanol in water as an example. This will also show you how to add explicit solvent molecules to your system:
Start ADFinput Build an ethanol molecule Pre-optimize the structure Use the panel bar Model → Solvation command Click on 'Explicit Solvent Molecules' Change the radius of the solvent sphere to a small value, such that 10 solvent molecules will be generated Click on 'Add Solvent'
As you can see, ADFinput has generated 10 water molecules around your ethanol molecule. It also has created two regions: a Solute region containing the ethanol molecule, and a Solvent region containing the water molecules. The visualization option for the Solvent region is such that the water molecules will only be shown using a Wire-frame representation.
The next step is to set up the QM/MM calculation:
ADF → QMMM: select the 'QMMM' panel Select the 'Solute' region in the 'QM Region' pull-down menu Change the QMMM Task to GeometryOptimization
Now you will have three tabs: the main QMMM tab that allows you to set QMMM details, the 'ADF 1' tab that is the setup for the ADF calculation for the Solute subsystem, and the 'MM 5' tab that sets up the MM calculation for the full system.
ADFinput is currently not very smart in setting the proper atom types for a MM calculation. So you will need to examine the atom types as they are generated in the MM input, and fix them if they are not correct. To fix this, if needed, use the Atom Inspector panel to set the Tripos (or Amber) atom types as needed.
Click on the 'MM 5' tab Activate the 'Run Script' panel (in the panel bar Details) Check the atom types (in the run script)
Run your calculation File → Run Show the optimization movie: In the ADFinput window: SCM → Movie Graph → Energy
Note that ADFmovie (and the other GUI modules) will only show the results for the QM part of your calculation.
To get detailed information on your QMMM calculation, you can check the output file using the SCM → Output command.
This tutorial will show you how to use the Fragment Analysis within ADF. The two examples used here are:
You will set up the calculations, run them, and visualize the results.
The resulting ADFinput input file (*.adf) is available in the examples directory included in the ADF distribution. In the examples documentation you can read how to set up a similar calculation without ADFinput.
The first example in this tutorial is Ni(CO)4. It consists of one Ni fragment and once CO fragment that is repeated four times.
The second example is PtCl4H22-. It consists of a PtCl42- fragment, and one H2 fragment. It is a good example on how to specify the charges of the fragments.
The structure is perfectly tetrahedral. You should be able to build the molecule yourself, using the techniques described in earlier tutorials. One possible way:
Start ADFinput Build a tetrahedral metal complex: Structure Tool → Metal Complexes → ML4 tetrahedral Change the central atom into a Ni atom Select one ligand Select all ligands: Select → Select Atoms Of Same Type menu command Change them in CO ligands: Atoms → Replace By Structure → Ligands → CO Choose the 'Geometry Optimization' Preset Save and Run When the run has finished, click 'Yes' to import the optimized coordinates Save
Your molecule should look something like this:
The fragments that ADF uses are based on the regions that you define. In this example we will generate four new regions: one for each of the CO ligands. The regions for the CO ligands will get special names to make sure that ADF recognizes them as one fragment repeated four times.
The Ni atom will not be in a region. ADFinput will automatically create atomic fragments for all atoms not in a region.
Repeated fragments are indicated with the fragment name followed by '/n', with n the number of the copy. All copies must match such that one fragment can be positioned exactly over another fragment by rotation and translation. ADF checks for this, ADFinput does not. In this particular example all four CO fragments are obviously identical by symmetry.
Panel bar Model → Regions Select all atoms Click the '+' button to add another new region (containing all atoms) Select the Ni atom Remove the Ni atom from Region_1: click the '-' button on the right side of the 'Region_1' Change the name of 'Region_1' into 'CO'
Now one new region has been defined: the CO region for the ligands. The next step is to split the ligand region into four repeated regions:
Press the triangle on the right side of the CO region line In the pop-up menu, use the 'Split By Molecule' command
This should result in the following regions:
The next step is very easy: we will tell ADF to perform a Single Point calculation (fragment analysis in ADF only works with a single point calculation), and we tell ADFinput to use the regions that we just defined as fragments:
Open the 'Main' panel Select the 'Single Point' preset Panel bar MultiLevel → Fragments Check the 'Use fragments' check box
Instead you could also select the 'Fragment Analysis' preset, which is identical to the Single Point preset but with 'Use fragments' activated.
In the Fragments panel you will see that one fragment is present, without charge: the repeated CO fragment. ADF will use basic atomic atoms for any atoms not put in another fragment. Thus, the Ni atom will be an atomic fragment.
Next you will Save and Run the calculation. When you do this, ADFinput will actually save two different calculations:
You do not need to set up a separate ADF calculation for the CO fragment: when you run the fragment analysis calculation, the run script will detect that the individual fragments first need to be calculated. It does so automatically. Next, the fragment analysis calculation will run.
Save Run Observe the running of the fragment (CO), and the final fragment analysis
The CO-fragment and Ni(CO)4 calculations all have been performed.
This calculation results in the normal .t21 and .out output files. You can view them with the SCM → View and SCM → Output commands. More interesting in the case of a fragment analysis is the interaction diagram that you can view using ADFlevels:
In the still open ADFinput window: use the SCM → Levels command
In the center you see the levels of the whole molecule, on the sides you see the CO fragment and the Ni-atom fragment. The interaction lines tell you which molecular orbitals come from which fragment orbitals.
In the output file you can find detailed information about the composition of the molecular levels in terms of the fragment orbitals.
The following is a picture of the PtCl4H22- molecule, after optimization with ADF::
To make this molecule, an easy way is to start with an octahedral complex to ensure symmetry. Next make changes to get this molecule, ending with a geometry optimization with ADF:
Build an octahedral metal complex using the Structures button (Structure Tool → Metal Complexes → ML6 octahedral) Change the central atom into a Pt atom Change four dummies in a plane into Cl atoms Remove one of the remaining dummies Change the remaining dummy in OH, via Structure Tool → Ligands → OH (this will ensure the final two H's do not break symmetry) Change the O atom into an H atom Choose the 'Geometry Optimization' Preset Set the total charge to -2 Select the Scalar ZORA relativistic option Run When the run has finished, click 'Yes' to import the optimized coordinates Save
Note that pre-optimization using UFF will make the geometry worse, if you wish you can pre-optimize using Mopac. But the ADF geometry optimization will also converge without pre-optimization.
Define the PtCl42- and H2 fragments in the Regions panel:
Panel bar Model → Regions Select the Pt and Cl atoms Use the '+' button to add a new region, and name it 'PtCl4' Select the two H atoms Use the '+' button to add a new region, and name it 'H2' Clear the atom selection (click in empty space)
Now the fragments are defined. Next, we set up the fragment analysis calculation:
Select 'Main' panel Select the 'Fragment Analysis' preset Panel bar MultiLevel → Fragments Change the charge of the PtCl4 fragment to -2 Save
The other changes (the overall molecule charge, and the scalar ZORA option) have already been set previously. The Fragment Analysis preset does not change these.
When you click on the Open button (the big dot) next to the PtCl4 fragment, you can inspect the PtCl4 fragment setup:
Click on the Open button (the big dot next to the PtCl4 fragment) in the 'Fragments' panel Check the charge of the fragment in the newly opened ADFinput (should be -2) Close the PtCl4 fragment ADFinput window
For more complex calculations, you could make additional changes to your fragment runs. However, normally the set up as created by ADFinput should be fine.
Next you will Run the calculation:
Run
After the calculation has finished, you can view the resulting interaction diagram:
Use the SCM → Levels command Use the mouse (drag with left mouse, scroll wheel, drag with right mouse) zoom in on the interesting region (roughly from -0.5 to 0.4) Select the PtCl4H2 column (by clicking on the name at the bottom Use the View → Interactions → Show menu command With the pop-up menu in the H2 column, shift the H2 levels by +0.28
Only interactions between visible levels are shown. So, if you zoom out no interactions will be visible for some of the levels. That is the reason that you will need to use the Show Interactions menu command.
We needed to shift the H2 levels to accommodate that in the final molecule the fragment experiences the -2 charge, but this is absent from the H2 fragment calculation. The interaction diagram should look something like the following:
In the output file you can find detailed information about the composition of the molecular levels in terms of the fragment orbitals.
This tutorial will help you to:
Start ADFinput Create an N atom Press Esc, and click in empty space so that nothing is selected. Create an H atom, not too far away Press Esc, and click in empty space so that nothing is selected. Add 2 more hydrogen atoms this way.
You should see one nitrogen and three hydrogen atoms in the drawing area now. We intentionally did not add bonds between the N and H atoms!
First we will try to optimize this structure using the default pre-optimizer (UFF):
Click 'Pre-Optimize' cog wheel button
Using the default pre-optimization algorithm will lead to a structure with way to big N-H distances:
Next we will try to optimize using Mopac.
Undo the previous optimization: Edit → Undo Right-click on the cog-wheel, select Mopac to pre-optimize
The message 'MOPAC optimization ready' will be displayed in the ADFinput window:
Mopac will produce a good-looking pyramidal ammonia molecule, with the N-H bond length of about 1.0 Angstrom.
Additionally, it will create proper N-H covalent bonding! These bonds are based on the bond-orders as calculated by Mopac.
Now with our pre-optimized NH3 molecule, we will set up a series of calculation: the optimization of NH3 with ADF, using different basis sets. The first step is to set up a simple geometry optimization:
Select the 'Main' panel Select the 'Geometry Optimization' preset Save as NH3
We have used all defaults for basis set and so on.
Next we will set up a series of calculations, using the Prepare tool.
Select the ADFjobs window In ADFjobs window, locate the 'NH3' job saved earlier Click onto 'NH3' line to select it Use the Tools → Prepare... command
The Prepare dialog window will pop up.
Because we have selected the NH3 job earlier, NH3.adf filename (including the path) will show up in the 'Run' field (otherwise, we can select NH3 using '+' button of the 'Run' field). This means we are going to create jobs that are identical to the NH3 job, but with some details changed (this will next be specified in the Prepare dialog).
Go to the 'Use these input options' and click '+'
Choose 'Basis Set and Frozen Core' → 'SZ' → 'Large'
In the similar manner, add basis sets 'DZ', 'DZP', 'TZP',
.. (optionally 'TZ2P' and 'QZ4P')
Use core 'Large' for all the basis sets
The basis set options selected will show up in the 'Use these input options' field. Note that by default the jobs will be created in the 'autojobs' directory, as specified in the 'Produce jobs' section.
Now click 'OK' in the Prepare window
The Prepare dialog will close and ADFjobs will open the 'autojobs' directory. Using NH3.adf, the Prepare tool created a set of jobs inside 'autojobs' named as 'NH3.SZ.Large.adf', 'NH3.DZ.Large.adf' etc. The files naming intuitively follows the basis set options which has been used, and the rest of the options of the newly created files are exactly the same as in NH3.adf. The original NH3.adf file has not been altered.
Now we are going to run the batch of ADF jobs which has been automatically prepared in the previous Step 3 of the tutorial Using ADFjobs, one can either run this set of jobs one by one, or all at once.
Click on the first job Shift-Click on the last job Run the jobs Job → Run
This will run the jobs in the default queue. Normally this is the Sequential queue, and the jobs will actually run one after another.
The jobs will start running, and the corresponding ADFtail windows will pop up. Each ammonia optimization job should not take more than a few minutes.
When all the ammonia optimization jobs are finished, we can compare the results. The main conclusion that you will reach is that DZ (double zeta) basis set optimization, job name 'NH3.DZ.Large', leads to the incorrect planar ammonia molecule. As covered in the preceding tutorials, you can optionally use several GUI modules to observe that:
- ADFmovie, which will show the geometry optimization
- ADFinput, which will open the (last) converged geometry
- ADFview, which will open the converged geometry and also would display various density maps in 3D
For example, let us use ADFview to demonstrate the results:
Clear the selection in ADFjobs (press Esc) Select job NH3.DZ.Large in the ADFjobs window (should then be highlighted) Select SCM → View Select Fields → Grid → Fine Select Properties → HOMO In the isosurface contour value field which appeared at the bottom, change the default '0.03' to '0.3' Type 'ctrl/cmd-minus' several times, until the central N atom ball does not overlap strongly with the HOMO orbital. Repeat for this for the DZP calculation
What you will observe is that HOMO of ammonia does not have the anticipated sp3 hybridization, when DZ basis set is used. In the following picture you can see both the DZ and DZP results:
To visualize results from several jobs automatically, you can use the Report tool within ADFjobs:
Go to the ADFjobs window Select Tools → New Report Template...
This will open up Report dialog.
In the 'Report' field, you can choose the name for your report file and its format ('HTML' or 'Tab separated plain text'). The other fields ('General', 'Images', 'Results') contain various relevant options you can choose for preparing your report.
In the 'Report'/'Template name' field, type 'MyReport' In the 'Results' section check 'Dipole Moment' and 'Dipole Vector' In the 'Extra ADFreport command line options' field: enter 'angle#labels#2#1#3
The extra command line options used instruct ADFreport to report the angle between atoms 2, 1 and 3, and include the atom labels for this angle on output. In a similar way more angles could be requested, or distances, or many other properties. For a description of all options please check the adfreport documentation.
Click 'OK'
The Report dialog will close, saving our template named 'MyReport'. Now we will generate actual report following the saved template:
In ADFjobs window, select all the finished jobs from the NH3 set Select Tools → Build MyReport Report
'Save As' dialog window will pop up, showing the 'report.html' default filename.
Choose your filename and location, then click 'Save'
Now ADF will work through all your selected jobs and prepare the report. Finally the browser will start showing your report file (by default, named 'report.html'):
The Report tool created a table-like report of the results. Click on the image above to see it in detail.
We can clearly see that only for DZ basis optimization our NH3 molecule is planar (check the last row of the report table). The dipole data are also shown in the report. Obviously the dipole moment is zero only for the planar NH3 structure.
Congratulations, we are done with the ammonia optimization tutorial!
If you want to exit all the GUI modules at once: select SCM → Close All
Start ADFinput
Next we need a reasonable guess for the structure of Caffeine. The quickest way to do this is to search for it in the database of molecules included with the ADF-GUI, and optimize it:
Press cmd-F or ctrl-F to activate the search box Type 'caffeine' in the search box Move your mouse pointer on top of the 'Thein' search result
As you can see, there are several matches. If you position your mouse over the results (without clicking) a balloon will appear showing the details of that match. For this tutorial we use the second match "Thein", from the NCI database. Thein is one of the common names for caffeine (and as you can see there are may alternative names),
Click on the 'Thein' search result
Click somewhere in empty space in the molecule drawing area to deselect the atoms
Switch to DFTB mode (panel bar ADF → DFTB)
Select SCC method SCC
Select the Dresden DFTB parameters:
click on the parameter folder icon, select Dresden
Note that only those parameter sets known to be able to handle your system will be shown in the menu.
If you move your mouse over the parameter field, the information balloon will also show references applicable to the selected set of DFTB parameters. More detailed information and references will be displayed if you click on the 'i' button next to the parameter input field.
Click the 'Run' button If the message says 'NOT converged', press 'Run' again.
The DFTB program should have created something similar to this structure:
Next we will calculate the AIM critical points and paths for the current structure.
Switch to ADF mode (panel bar DFTB → ADF)
Now we want to activate the Bader AIM analysis to find the critical points and bond paths. To find where this option is located, search for it:
Activate the search box (cmd/ctrl-F) Type 'criti' in the search box Use the Return key to accept the highlighted match (Other...)
ADFinput will activate the panel that displays the option you are looking for (to calculate the AIM critical points and paths). The matching input options will be marked with blue italic text. Note that we first had to activate the ADF mode, the input option search will restrict the search to panels that belong to the current method (ADF, BAND, DFTB, ...)
Check the box to calculate Bader (AIM ) Critical points, bond paths and atomic properties Check the box to calculate AIM atomic energies Check the box to save the Bader basins
Run this setup: File → Run
A dialog will pop up in which you must specify a filename to use for your job, for example caffeine:
Enter 'caffeine' as a Filename, pres the Save button
After hitting the save button the calculation will start. You will get two extra windows: first a window for ADFjobs that allows you to manage your jobs and keep track of their state (for example, queued or running). You will also get a window showing the ADF log file. This shows you what is going on in the current calculation.
Depending on your computer, the calculation should be ready after a few minutes at most:
Start ADFview SCM → View Show the HOMO Properties → HOMO
Hide the HOMO by unchecking the check box at the lower left corner of the ADFview window Add → Isosurface: Colored In the first field selector (to the right of the 'Isosurface: Colored' text at the bottom), select Density → SCF In the second field selector (to the right of the '0.03' text in the same line), select Potential → SCF
Hide the surface with the potential energy: uncheck the check box at the lower left corner of the window Properties → AIM (Bader)
The critical points and bond paths are shown (the molecule balls and sticks representation is hidden). The different types of critical points (atom CP, bond CP, ring CP and cage CP) are indicated by different colors. The atom CPs are scaled by density by default, which makes them look like atoms. The bond paths are colored by density, by default.
You can also visualize the Hessian of the Density in the critical points:
Uncheck the 'Scale By Density' check box in the AIM line at the bottom of the window Properties → AIM: Hessian of Density at CPs
To get a rough display of the Bader basins, use the Bader sampling option:
Properties → Bader Sampling Zoom in
The different colored points show the different basins.
ADFview has many options to visualize the results, the options just used are mainly to show off some features. Play around with the different options, for example try out what the check boxes do on the left side. Or try other fields, or colored cut planes, or ...
This finishes the Caffeine Bader (AIM) tutorial, close all its windows:
SCM → Close
Start ADFinput Make a benzene molecule (for example by searching for it with cmd/ctrl-F) Set up a Single Point calculation without frozen cores Panel bar Properties → Other: Etot, Bader, Charge Transport Check the 'Bader Atomic energies' option Check the 'Save Bader atomic basins' option Panel bar Properties → Localized Orbitals, NBO Check the 'Perform NBO analysis' option Request Boys-Foster localized orbitals Run this setup (File → Run)
When the calculation is done (it should run very fast), we use ADFview to examine the Bader charges and compare them with Mulliken charges:
Open the results with ADFview Show the Bader atomic charges (Properties → Atom Info → Bader Charge → Show) Color the atoms by Bader charges (Properties → Color Atoms By → Bader Charge) Show the Mulliken charges (Properties → Atom Info → Mulliken Charge → Show)
Next we inspect the NBOs and Boys-Foster localized orbitals. To remove the charge display we close and open ADFview, but you could also have used the View menu to remove them by hand:
Close ADFview Open the results again with ADFview Add a Double Isosurface Use the field menu in the new control line, and observe the labels present with the NBOs and NLMOs Open NBO number 15 (should be similar to C6-H12) Improve the grid by using Fields → Grid → Fine
Obviously, you can also visualize the NLMOs or the Boys-Foster localized orbitals (which are just called Localized Orbitals in the fields menu.
Next we inspect the Bader atomic basins. The numerical integration points are used for this purpose. The color indicates to which atomic basin the numerical integration point belongs to.
Close ADFview Open the results again with ADFview Properties → Bader Sampling
One can also select one or more atoms, to see only the Bader atomic basins of the selected atoms.
Now we will remove one electron from a selected orbital (by symmetry) using ADFinput again.
Open your benzene calculation in ADFinput Model → Spin and Occupation
In the Spin and Occupation panel you see which orbitals are available (by symmetry, and spin if the current set up is unrestricted). The numbers are the occupations. Next to the ADFinput window ADFlevels will also have opened, and is showing a level diagram of your current molecule. This helps you to select the proper occupation.
The level diagram is based on the existing result file (.t21) found. If none, ADFinput will suggest to do a guess-calculation: a hopefully fast calculation using an inaccurate integration grid, and stopping after a few SCF cycles. The level diagram shown in that case will not be accurate obviously but will still be helpful in selecting the proper occupation.
Now, just as an example, we make a hole in the 3E2.g orbital. Note that this is NOT the HOMO.
Change the 4.0 to 3.0 in the 3 E2.g column (the third row)
Close the guess adflevels File → Save Close the ADFinput window Run (via the Job → Run command in ADFjobs When the run is complete: Open ADFlevels using the SCM → Levels command In the ADFlevels window, use the View → Labels → Show command Zoom in to the region of interest (for example using your mouse wheel)
In the levels diagram you will observe that the HOMO (the 1E1.g) is completely filled, while the orbital below it has a hole. You can easily see this by the color. Note that we have performed a restricted calculation, so in effect the hole is averaged over the E orbitals.
This tutorial will help you to:
Embedded into proteins and coordinated by cysteine ligands, iron-sulfur cubanes are often used by nature in electron transfer and catalysis. While their native structures can be extracted from protein data bank (PDB) files, we will use ADFinput module to build a model of the Fe4S4 cubane from scratch in this example.
Start ADFinput Select the Structure tool → Polyhedra → Cube structure: Click anywhere in the empty structure drawing area. This should place a cube in the building area of ADFinput.
The cube built is constructed of carbon atoms. We will now change four of these atoms to iron (Fe), and next the remaining four atoms into sulfur (S).
Via Shift-click, select four carbon atoms in the corners of the cubane
Atoms → Change Atom Type → Fe: change the four atoms into Fe Selection → Invert Selection Atoms → Change Atom Type → S: change the other atoms into S
Now you should see the Fe4S4 cubane in the structure drawing area of ADFinput. The proper coordination to Fe atoms is important in modeling their electronic structure. In proteins, iron-sulfur cubanes are coordinated by cysteine ligands to the Fe sites. Here, we will model these four cysteines by -SH ligands. The procedure to add these ligands is described below.
Select all Fe atoms (for example by inverting the selection) Use the Atoms → Details (Color, Radius, Mass, ...) menu command Change the number of connectors for all Fe atoms from 10 into 4
Switch to the Main panel Add hydrogens to the still selected Fe atoms using the Atoms → Add Hydrogen menu command Select the (newly added) hydrogens Use the Atoms → Replace By Structure → Ligands → OH menu command:
This will replace hydrogen atoms into OH ligands.
Select one of the O atoms Use the Select → Select Atoms Of Same Type command to select them all Use the Atoms → Change Atom Type → S to change them into S
The next step is to optimize this structure. It is a difficult system, and the pre-optimizers will fail. So we will use ADF to optimize the geometry.
Set the charge to -2 Select the Geometry Optimization preset Click the Symmetrize button (the star) to check the symmetry (should be Td)
File → Run (enter FeS as name) When ready, click OK to use the optimized geometry in ADFinput SCM → Movie In the Movie window: Graph → Energy
Close the movie window: File → Quit
In Fe4S4 systems, the iron sites are commonly high-spin ferrous (Fe3+, S = 5/2) or ferric (Fe2+, S = 2). For the present example, we will use the iron-sulfur cubane oxidation state where the two sites are ferric and the remaining two are ferrous. This oxidation level of Fe4S4 is well-defined and occurs, for example, in rubredoxin and high-potential iron-sulfur proteins (HIPIPs). For our model system, Fe4S4(SH)4, this implies the total charge of -2.
The relative directions, or coupling, of the individual site spin vectors is a very important issue in obtaining the desired density functional solution in Fe4S4, as well as many other systems which display a multi-center radical character.
Within the common open-shell approach, the spin vectors are either parallel or anti-parallel. The case when all the spins are parallel is called high-spin (HS). Obtaining self-consisted field (SCF) solution for the HS case is normally simpler because the program does not need to resolve the ambiguity of distribution the sites spin vectors. While the ferromagnetic HS solution commonly does not correspond to the lowest energy electronic state, this solution can be used to obtain the electron density corresponding to the lower energy spin-coupled state. In this step, we will obtain the HS solution for the iron-sulfur cubane, which will be used later in the tutorial. The HS solution for the [Fe4S4(SH)4]2- model corresponds to S = 2 × 5/2 + 2 × 2 = 9.
Select the ADFinput window Select the Single Point preset Keep the total charge to -2 Set the spin polarization to 18 (corresponds to S = 9) Check the unrestricted box File → Save As... Enter FeS_HS as filename and save File → Run
ADFtail window will pop up showing the progress of the job:
The logfile shows that the convergence has been obtained in 17 SCF cycles.
While SCF solution for the high-spin (HS) state of a multi-center spin system can often be easily found, this solution does not necessarily correspond to the lowest energy state.
To generate the solution with the desired collinear spin arrangement, one option is to use the 'SpinFlip' concept that has been earlier introduced by L. Noodleman and coworkers. In this two step procedure:
Because this approach often results in lowering the electronic symmetry of the system while retaining its structural symmetry, a solution obtained in such way is often called the broken symmetry (BS) state. In such cases you will need to make sure that your BS calculation is done with lower symmetry.
The concept of SpinFlip and BS state can be illustrated considering our iron-sulfur cubane case with two ferrous (Fe3+, S = 5/2) and two ferric (Fe2+, S = 2) sites. One of the well-characterized BS states for this level of Fe4S4 oxidation corresponds to S = (5/2 + 2) - (5/2 + 2) = 0, 2↑:2↓.
Select the ADFinput window with your FeS_HS calculation Make sure the 'Main' panel is visible Change the spin polarization from 18 to 0
This spin polarization setting corresponds to S = 0 zero spin of the BS electronic state.
Panel bar Model → Spin and Occupation
ADFlevels will show the levels from your previous calculation. As we are not changing the occupations this info is not needed for this tutorial.
Close the window showing the energy level diagram (ADFlevels ...) Select the two (out of four) arbitrary Fe sites in the drawing area Click + next to the Spin Flip on Restart For: line.
To achieve the desired BS solution, the SpinFlip procedure should be applied to 2 out of 4 Fe sites of Fe4S4. In the above example, we selected sites Fe(2) and Fe(7). This will instruct SpinFlip algorithm to interchange α (↑) and β (↓) electron densities associated with these two Fe sites when the job will be restarted, changing the HS 4↑:0↓ spin state to the BS 2↑:2↓ spin state.
The Spin Flip option only works when restarting, so set up the restart calculation from the HS results:
Panel bar Details → Files (Restart) Click folder icon in front of the 'Restart file:' field, Select the Fe_HS.t21 file (it should already be selected, so just press the Open button)
The above will instruct ADF to read the converged HS solution we obtained in the previous step of the tutorial. This solution has been saved in the Fe_HS.t21.
Panel bar Details → Symmetry Set the symmetry to NOSYM File → Save As, save the job as FeS_BS_SPINFLIP File → Run
As you see, our FeS_BS_SPINFLIP job converges in 21 cycles.
There is an alternative to SpinFlip available in ADF, which is aimed to achieve a specific spin-coupled solution in a single calculation only. This is done using the MODIFYSTARTPOTENTIAL key in ADF: it allows you to create a spin-polarized potential at the very start of the calculation.
Please read the page of ADF user's guide on MODIFYSTARTPOTENTIAL key prior to proceeding with this step of the tutorial. As follows from the MODIFYSTARTPOTENTIAL description, this key allows to control the ratio of spin-α and spin-β electrons associated with fragments via 'alpha' and 'beta' numbers. For the purpose of the present tutorial, we will consider the four Fe sites as fragments. Apparently, 'alpha' and 'beta' numbers will correspond to the number of spin-α and spin-β electrons, correspondingly, associated with a Fe site.
So follow these instructions to obtain the BS solution via the MODIFYSTARTPOPTENTIAL option:
Open your FeS_HS calculation in ADFinput
Change the spin polarization from 18 to 0
Panel bar Details → Symmetry
Set the symmetry to NOSYM
Panel bar Model → Spin and Occupations
Close the ADFlevels window
Select the four Fe atoms
Click the '+' in front of 'Modify Start Potential'
Set the alpha and beta occupations, for the four Fe atoms:
14 alpha and 10 beta
14 alpha and 9 beta
10 alpha and 14 beta
9 alpha and 14 beta
For the spin-up Fe3+, S = 2, the alpha and beta numbers would be 14 and 9, correspondingly, and for the spin-up Fe2+, S = 5/2, these numbers are 14 and 10. For the spin-down Fe sites, alpha and beta numbers should be apparently transposed. Also, our desired BS state for this level of Fe4S4 oxidation corresponds to S = (5/2 + 2) - (5/2 + 2) = 0, 2↑:2↓ .
Note that SCF procedure might be problematic.
You can play with several options to help the convergence. In the SCF panel (in Details) you can experiment with several mehods like LISTi, A-DIIS, E-DIIS and ARH. In the SCF convergence details panel further options are available, like mixing, level shifting, orbital freezing and DIIS details.
In this particular case, the ARH method (in the SCF panel) works, although it needs a lot of iterations.
Panel bar Details → SCF Select the ARH convergence method Set the maximum number of SCF cycles to 1000
File → Save As: save the job as FeS_BS_MODIFYSTARTPOTENTIAL File → Run
Hopefully you will be able to converge the job, to the same energy and state as the SpinFlip job.
Thus both the SpinFlip and the ModifyStartPotential option allow you to obtain a desired Fe spin coupling pattern in the Fe4S4 case. While SpinFlip is a two-step approach and ModifyStartPotential works as in a single step, the SpinFlip approach shows a better performance during SCF (much better and faster SCF convergence).
In the two previous steps of this tutorial, we have generated the broken symmetry (BS) solution for the Fe4S4 cubane using alternatively the SpinFlip and ModifyStartPotential options. Here, we will analyze this BS solution viewing the Fe spin densities using ADFview, and confirm that the spin alignment of the iron sites is 2↑:2↓ This type of analysis can also be powerfully used during presentations and for scientific illustrations.
Select your FeS_BS_SPINFLIP calculation in ADFjobs Use the SCM → View menu command to activate ADFview
You should see the [Fe4S4(SH)4]2- system in the ADFview window.
The next step is to calculate the spin polarization density:
Fields → Calculated In the Filed:Calculated line, select the Density A → SCF field to the left of the '-' Next, select the Density B → SCF field to the right of the '-'
The minus '-' sign separating the tabs implies that we have requested ADFview to calculate the 'spin-α - spin-β' density field. ADFview automatically assigns 'C-1' title to this field, see left to the Field:Calculated.
We will now visualize this 'C-1' field using one of the available ADFview options:
Add → Isosurface: Double (+/-) In the new control line, select the Other → 1 → C-1 field
You should see now the two Fe ions surrounded by blue blobs (spin up,↑) and the other two by red blobs (spin down,↓):
In the same way you can also check the spin densities for the FeS_HS and FeS_BS_MODIFYSTARTPOTENTIAL results.
This section contains tutorials that show you how to use BAND via the BAND-GUI. As the BAND-GUI is very similar to the ADF-GUI, we will focus here on BAND specific features.
If you are not familiar with the ADF-GUI (or the BAND-GUI), please have a look at the Introduction and ADF-GUI Overview tutorials.
For example, it will show you how to start the BAND-GUI.
The Building Crystals and Slabs tutorial in the overview section is especially important for BAND users.
According to any freshmen chemistry textbook, in NaCl one electron is transferred from the Sodium to the Chlorine. The occupied 3p states form the valence band, while the empty sodium states hybridize into a conduction band. We will put these idealized ideas to the test.
This tutorial will teach you how to:
The BAND-GUI has been designed to be a lot like the ADF-GUI. This makes it much easier for users to use both programs. To avoid repetition, the BAND-GUI tutorial assumes that you are familiar with some basic usage of the ADF-GUI. If you do not know how to rotate, translate, zoom etc within the ADF-GUI, please read through the first ADF-GUI tutorial before starting with this BAND-GUI tutorial. Even better: try using the ADF-GUI yourself. You can get a demo-license for this purpose if needed.
Start ADFjobs
We prefer to run the tutorial in a new, clean, directory. That way we will not interfere with other projects. ADFjobs not only manages your jobs, but also has some file management options. In this case we use ADFjobs to make the new directory:
Select the File → New Directory command (thus, the New Directory command from the File menu) Rename the new directory by typing 'BandTutorial' and a Return Change into that directory by clicking once on the folder icon in front of it
Now we can start BANDinput using the SCM menu:
Use the SCM → BANDinput menu command
The BANDinput window consists of the following main parts:
In the drawing area you see the lattice vectors, with three repetitions of the unit cell along each vector. The first lattice vector is red, the second green, and the third blue. Using the View Periodic menu you can change the display of the periodic images:
View → Periodic → Repeat Unit Cells
If you look in the toolbar with molecular editor tools, you see that the rightmost button is no longer glowing. You can also use this button to toggle the display of the periodic images:
Click once on the rightmost button in the toolbar (the four dots)
Now you should see the periodic repetitions again.
Salt has an fcc lattice. First we need to set the lattice vectors:
Panel bar Model → Lattice Enter the lattice vectors, as shown below
Now we will add the Na and Cl atoms. It is convenient to turn of the periodic display while doing this:
Uncheck View → Periodic → Repeat Unit Cells Click on the X tool, and select Na
After this you see at the bottom of the screen "Na tool" in the status field:
Click once in the drawing area, near the origin Click once on the created atom to stop bonding
As you can see the atom is not exactly in the origin. This can be fixed if you wish:
Edit → Set Origin
To add the Cl atom you can proceed the same way:
Select the Cl tool (via the X button again) Click once somewhere in the unit cell Click on the new Cl atom to stop bonding
Next you should edit the Cl coordinates and change the Cl color:
Select the coordinates panel: panel bar Model → Coordinates Change the Cl coordinates to be (2.75,2.75,2.75) Check View → Periodic → Repeat Unit Cells
Now your system looks like:
If you wish so, you can give your calculation a title:
Panel bar Main Enter NaCl in the "Title" field File → Save, name it "NaCl" and press Save. File → Run
A window will appear showing the progress of the BAND calculation (the 'logfile'). After a few minutes the calculation has finished, and it looks like:
Wait for the text 'Job NaCl has finished' in the logfile
The calculation has produced two important result files: "NaCl.out", which contains the result of the calculation in text format, the second is"NaCl.runkf" which is a binary result file.
SCM → BAND Structure
This will open the bandstructure window:
It consists of a plot and a picture of the Brillouin zone. In the plot the red line is the Fermi level. Below the Fermi level are four occupied bands. You can see this more clearly by vertical zooming:
Click on the right mouse button, and drag the pointer up to zoom vertically When the region of interest gets out of view, drag it into view (with the left mouse button)
The bottom part of the plot will look like:
In most k-points you see now four bands below the Fermi level. In some k-points you see fewer because they are degenerate.
Now you may wonder about the path taken through the Brillouin zone. This is where the bottom part comes in handy. You can zoom and rotate, just as with molecules in BANDinput.
Zoom in a bit, by holding the right mouse button and moving upwards
Toggle the button to display the k-points
Now click on the line from 11 (via 12) to 1
Note how the line lights up, and also how the corresponding segment is indicated in the plot by a gray background. You can also click on the plot to select line segments.
Rotate the Brillouin zone a bit to convince yourself that the line (from k-point 11 to 1) runs from the center to the center of a hexagonal face.
Now what is the character of the bands? Let us first examine this narrow band at about -0.5 Hartree.
SCM → View In ADFview: Add → Isosurface Double (+/-)
In the bar at the bottom of the window, you can select which field to show.
Select the lowest band (k=0,0,0)
From the label you can see that it has an energy of -0.8909 and the coordinates are (0,0,0).
A progress bar will appear, and after a while you will see the orbital:
If you rotate it a bit and toggle the isosurface on and off (with the check box in front of it), you can convince yourself that this orbital is located around the small atom, which is the Chlorine.
Toggle the periodic view (the menu View → Periodic → Repeat Unit Cells)
Obviously this is the 3s band of Cl. The strange truncated spheres are due to contributions of neighboring cells.
Let us now take a look at the orbital with the lowest energy of the second band (the first one with an energy not around -0.88):
Select the lowest orbital of the second band (with energy about -0.05)
This orbital looks like:
and it clearly consists of a p orbital on the Cl. The part near the Na comes from Chlorine atoms in neighboring cells.
It is generally easier to interpret orbitals at k=(0,0,0). Going up in energy we encounter a degenerate triple at e=-0.49. Note you can tear of the menus to make orbital selection easier:
Tear off the 'Occupied orbital menu, group 1 Select the first orbital around e=-0.49
Take a look at all three of them.
From these orbital pictures we can conclude that the valence band is indeed mainly of Chlorine-p character.
You may want to check the lowest orbitals of the (unoccupied) conduction band.
Check the lowest orbitals of the conduction band Do you see a strong Na character in them?
Actually it looks a lot like an s function on the Cl atom. However in this case the iso-surface is misleading. With a contour plot you can make it look like this
so the orbital has a bit of a muffin-tin shape, the maximum is between the atoms. For such states an iso surface looks misleading. The maximum is closer to the Na atom, so it is really a Na state.
There is in fact a much more easy way to conclude that the valence band is mainly of Chlorine-p character.
Open the DOS module: SCM → Dos
and a window like this will appear:
The Fermi energy is around -0.43, and there is clearly a gap. Just below it there is a valence band, and at -0.2 Hartree starts the conduction band.
Select the View → Add Graph command
(Now you see two plots of the total DOS)
Select with the mouse the Chlorine atom (the small green one)
It is already immediately clear that the valence band comes from the Chlorine, and the conduction band from the Sodium.
Right-click with the mouse on the selected Cl atom and check the 'P-DOS' check box in the pop up menu.
This shows that the valence band is clearly made of Chlorine p-orbitals.
Naively you might expect that the charge of Na should be +1 and from the Cl -1. This can be best seen by making a cut plane:
Go to ADFview (SCM → View) Delete the double isosurface with the 'Isosurface: Double' pull-down menu in the bar on the bottom and select 'Delete' Add → Cut Plane: Contours (+/-) Select the deformation density in the fields menu (in the Fit density sub-menu)
The result is a bit disappointing, because the plane is perpendicular to the atoms
To get it right enable at the bottom the "Position plane" toggle:
You will now see a white frame. Drag it to the right until you see an arrow in it
Now you can "grab" the arrow head and turn it to make the plane go through the atoms
Indeed we see that charge is added (blue) near the Cl and removed (red) from the Na atom. The trend is good, but what is the total amount of charge transferred?
Use the "Properties → Atom Info → Hirshfeld Charges → Show menu command
This will show you
So the amount of charge transferred is only about 0.2. This is of course due to the fact that the Cl p-band overlaps quite significantly with the Na region.
The conclusion of this tutorial: we should take the idea that one electron is transferred from Na to Cl with a grain of salt.
This tutorial will teach you how to:
Throughout we will consider the toy system of a periodic chain with three atoms in the unit cell.
We are going to enter the geometry manually, just as in the first tutorial.
Start BANDinput Set periodicity to Chain Panel bar Model → Lattice Set the lattice constant to 10.0
Add with the mouse three hydrogen atoms somewhere in the cell Panel bar Model → Coordinates Change the coordinates of the H atoms to match the following picture
You have now created the cylinder symmetric toy system.
Go back to the 'BAND Main' panel Choose as preset 'Geometry Optimization' File → Run, save the job with name H3_geo
After it has finished the program asks you
Answer 'Yes'
Now let us look at the progress as a movie.
SCM → Movie
A small movie pops up ending like this
The geometry looks a bit odd. Now take a look how the energy changed during the optimization.
Choose the Graph → Energy menu command
It shows the energy at the four steps: 0, 1, 2, and 3. Since the energy does not change anymore from step 2 to three it should be OK? Well, maybe, but maybe not. Let us check whether we were fully converged.
Open the logfile: SCM → Logfile Scroll from the end a bit upwards and you will see
You see the final geometry and status of the five convergence criteria. Because they are all satisfied you see the log message 'Geometry Converged'
Maybe we are dealing with a shallow minimum. Let us retry with a more strict criterion.
Close the logfile and movie windows and go back to 'BANDinput' Go to the 'Geometry Optimization' details panel (click on the "..." button next to GeometryOptimization) Set the gradient convergence criterion to 0.0001
File → Run (click Yes twice to accept overwriting the previous job After the optimization, click Yes to accept the new geometry in BANDinput Open ADFmovie and show the energy
Now that looks more like an optimal geometry!
Go to BANDinput Select in the 'BAND Main' panel the 'Frequencies' preset
File → Save As... Save the project as H3_freq File → Run When it is done, click Yes to accept the new coordinates
After the calculation has finished
SCM → Spectra
There appear to be three peaks, whereas you would expect 3N degrees of freedom. With three atoms (N=3) we should have nine modes. We can examine this a bit closer
click on the 'NormalMode' menu.
So there are indeed nine vibrational modes. Only two are nonzero because only symmetrical modes are calculated by default. To see what a mode looks like
Select the mode at 1662 cm-1, either from the 'NormalMode' menu, or by clicking on it directly in the graph.
A new movie window pops up visualizing the vibrational mode.
A minimum has vanishing gradients and only positive eigen modes. A transition state (first-order saddle point) is characterized by having one negative mode. With a transition state search the optimizer will go uphill in the direction of the lowest (nonzero) eigen-mode and downhill in all other degrees of freedom. In our example it would follow mode 8. Let us give it a try from the minimum.
Choose in the 'BAND Main' panel the preset 'Transition State Search'
We have just calculated a Hessian (with the frequency run) so we'd better use it.
Use the panel bar Details → Geometry Convergence command Click on the folder button next to 'Initial Hessian From:' Select with the file dialog 'H3_freq.runkf'
File → Save As, use name H3_ts File → Run After it has finished: Update the coordinates in BANDinput SCM → Movie Graph → Energy
The third H atom ends up exactly in the middle of the (repeated) H1 and H2 atoms. Let us finally check that we are indeed in the transition state.
In the 'BAND Main' panel select the task 'Frequencies' Save the project as 'H3_ts_freq' and run it. When the calculation is ready: open ADFspectra and click on the 'NormalModes' menu
You should see:
We have found a geometry with vanishing gradients with one weak negative vibrational mode. We have succeeded in finding a transition state.
SCM → Close All
This tutorial will teach you how to:
In this "advanced" tutorial we consider a slightly more realistic system. Some of the calculations may require 20 minutes to run on a two core machine.
We are going to make a one layer Li (001) slab with a 2x2 unit cell, assuming familiarity with the build tools
Start BANDinput From the structure tool select 'Cubic' and 'bcc' Set 'Element' to Li Set the lattice parameter to 3.49 Press 'OK' Invoke the Slice tool Set the Miller indices to 001, select 'Cartesian', and enter 1 layer. Press 'OK' Use the Edit → Crystal → Generate Super Cell... command Select the preset '2x2' and press 'OK' Panel bar Model → Coordinates
Your screen should look like this:
Add with the mouse two hydrogen atoms anywhere in the screen Set in the table the coordinates of the first hydrogen atom to (0, -0.5, 2) Set the second H atom coordinates to (0, 0.5, 2)
The final geometry looks like this
Select the 'BAND Main' panel Select the Frequencies preset Set 'Basis Set' to 'SZ'
Go to the 'Frequencies' details panel (click on the '...' button) Select with the mouse the two tiny Hydrogen atoms (zoom in to make selecting them easier) Click on the '+' button next to 'Partial Hessian For:'
File → Run, name the job H2onLi_freq Say 'No' when asked to update the coordinates
Let us examine the eigen-modes that we have found for the Hydrogen molecule
Select the SCM → Spectra menu command Open the 'NormalMode' menu
Now you will see that there is an eigen-mode at 452 cm-1 and one at 2422 cm-1. Convince yourself that the 452 mode moves the H2 perpendicular to the service and that the 2422 mode is essentially an H2 stretch mode. The lowest mode looks like a promising start to find the transition state for dissociation over the Li surface.
We have just found the vibrational modes of the Hydrogen molecule, assuming that the Li substrate remains fixed. Let us now find the transition state under the same assumption.
Close the 'ADFspectra' window and go back to BANDinput. Select the 'BAND Main' panel and select the 'Transition State Search' preset
Panel bar Model → Geometry Constraints and Scan The two hydrogen atoms are still selected, if not select them again Use the Select → Invert Selection command Click on the '+' button next to 'freeze selected atoms'
Panel bar Details → Geometry Convergence Set the initial Hessian to 'H2onLi_freq.runkf'
File → Save As..., use name H2onLi_ts File → Run
After the run has completed:
SCM → Movie In the movie window: Graph → Energy View → View Direction → Along x axis Reorient to get a good look
It should look like
DFTB provides accurate results at a fraction of the cost of a DFT evaluation. The tutorials in this section show you how to use DFTB via the DFTB-GUI. Since the DFTB-GUI is very similar to the ADF-GUI, we will focus here on DFTB specific features.
If you are not familiar with the ADF-GUI (or the DFTB-GUI), please have a look at the Introduction and ADF-GUI Overview tutorials.
The DFTB-GUI can be launched from ADFJobs or by changing the program from the pull-down menu to DFTB. As such, one can switch seamlessly between different programs, for instance to use DFTB as a pre-optimizer for an ADF or BAND calculations.
This tutorial will walk through the main features currently provided by the DFTB engine, using the GUI. DFTB achieves evaluation accuracy at a much lower computational cost than DFT, making it suitable for molecular dynamics evaluations, even on large systems.
We will walk through three different steps in this tutorial. The first step aims at pre-optimizing a simple molecule with DFTB. The second step will evaluate IR frequencies, spectrum and vibrational modes. Finally, the third and last step will perform a simple molecular dynamics evaluation.
Start ADFjobs SCM → DFTBinput
Note that DFTBinput and ADFinput are actually identical. By starting DFTBinput from the SCM menu ADFinput will immediately switch to DFTB mode.
In the ADFinput panel: cmd/ctrl-F, find benzene Select the entry Benzene(ADF) in the Molecule section of the search result Select the Main panel Make sure the GeometryOptimization task is selected Make sure the SCC method is None Make sure the parameter set is "Dresden" Move your mouse over the parameter set option, and note the references in the help balloon Click 'Run' in the DFTB panel to optimize the molecule. When done (DFTB Ready in the lower left corner of the window): Use the View → Atom Info → Atomic charge: Net (dftb) → Show menu command to show the DFTB charges on each atom
A Self Consistent Charge (SCC) evaluation allows atomic Mulliken charges to vary in an iterative procedure. This influences the DFTB Hamiltonian until self consistence of these charges is reached.
Enabling SCC requires higher computational cost due to the iterative procedure taking place for each energy evaluation, but higher accuracy of the final result. If SCC is disabled, the resulting final charges are not self consistent, thus the procedure will be faster, but less accurate.
Set the SCC method to SCC Click 'Run' again in the DFTB panel to optimize the molecule.
Note how both equilibrium geometry and atomic charges are affected by the SCC evaluation:
This example will show how to compute vibrational frequencies and modes using DFTB.
Choose Frequencies as the Task to perform.
File → Run, use DFTBFrequencies as name when asked for a name When the calculation is ready (Job DFTBFrequencies has finished in the logfile): Close the logfile (File → Quit) In ADFjobs: Select the "DFTBFrequencies" job SCM → Spectra Click on each individual frequency band to show the associated vibrational mode.
In the ADFinput window, choose MolecularDynamics as the Task to perform. Make sure the SCC method SCC option is selected, and the parameter set is Dresden Click on the Details button (...) to the right of the Molecular Dynamics task Set the Number of steps to 2000, the Time step to 0.5, and the Sampling frequency to 5 Make sure the thermostat is set to Berendsen, and adjust the temperature to 500.0 kelvin. Save your input choosing File → Save as ..., with name "DFTB-MD"
File → run
You can monitor the calculation in real-time:
While the calculation is running: SCM → Movie In the ADFmovie window: Graph → Energy
This tutorial will walk through the use of periodic DFTB to optimize the lattice parameters. The density of states and band structure of the optimized structures will be displayed.
To perform geometry optimization of a lattice, we will start from a diamond crystal with incorrect lattice values, and let DFTB optimize this value. Next, we will proceed to display additional information obtained during the evaluation.
Start ADFinput Select the DFTB panel: panel bar ADF → DFTB Choose Bulk to enable periodicity. Click the structure tool (the snowflake next to the X tool), Choose Cubic → Diamond Set the cell parameter a to 6.0 and click OK
Click the Details button (...) for the GeometryOptimization task Check the Optimize Lattice box Go to the Main panel Make sure SCC is set to None, and that the Dresden parameter set is selected Choose File → Save as... and save the input as "DFTB-Diamond".
File → Run. While the optimization is running, you can watch the progress by clicking SCM → Movie Update the geometry in ADFinput when the job is finished
From the movie, you can observe how the unit cell shrinks as it is optimized.
You can visualize the density of states (DOS) directly from the converged optimization.
SCM → DOS (in the ADFmovie or ADFinput window)
Similarly, you can visualize the band structure
SCM → Band Structure
Activate the ADFinput window with the DFTB-Diamond input Select "Phonons" as Task File → Save as..., use "DFTB-Phonons" as filename File → Run When completed: use SCM → Band Structure in the ADFtail window (showing the logfile)
A window will appear showing the phonons diagram. Clicking on the curves at the location of the vertical red bars will show the actual crystal vibrations.
In this tutorial we will calculate a proton affinity, defined as the negative gas phase enthalpy for the protonation reaction A- + H+ → AH. Gaus, Cui and Elstner have shown that third-order terms (DFTB3) generally improve the proton affinities with respect to second order self-consistent charges (SCC).
We will calculate the proton affinity of the acetate anion, CH3COO-. In the first step of this tutorial, we will optimize acetic acid at the DFTB3 level. The second step will perform the computation on the anionic species, and compute its proton affinity.
Contact us to enable parameter files from DFTB.org, including the ThirdOrder set for DFTB3, needed to run this tutorial. For non-profit users there is no additional charge for this.
Start ADFjobs SCM → DFTBinput
Use the Structure tool, select Solvents → Acetic acid Click anywhere in the visualization pane. The acetic acid molecule will appear. Make sure the GeometryOptimization task is selected and charge is set to 0.0 Select DFTB3 as the SCC method Select the ThirdOrder parameter set. Click 'Run' in the DFTB panel to optimize the molecule.
When the optimization is complete (DFTB Ready in the lower left corner of the window), note the final energy (for the paramenter set used in this example, E(HAc) = -11.610899 Hartree)
Select the hydrogen in the COOH group by clicking on it. Press the Backspace key on your keyboard. This will delete the hydrogen atom. Select the Main panel Set the charge to -1.0 Click 'Run' in the DFTB panel to optimize the molecule.
At the end of the procedure, note the final energy (for the parameter set used in this example, E(Ac-) = -11.280518 Hartree).
Additionally, the energy of the lone proton must be obtained.
Select and delete every atom of the molecule, except one hydrogen atom. Select the Main panel Make sure the GeometryOptimization task is selected. Although this is effectively a single point calculation, the GeometryOptimization task reports the energy directly in the molecule window, facilitating the read out of the energy.
Set the charge to 1.0 Click 'Run' in the DFTB panel.
Finally, annotate the value for E(H+) = 0.236417 Hartree. The proton affinity is computed as PA = E(Ac-) + E(H+) - E(HAc), resulting in a final Proton Affinity of 0.566798 Hartree, or 355.67 kcal/mol.
We leave it as an example to calculate the PA with DFTB2 (SCC), and compare it with the high-level ab initio results that are also quoted in the DFTB3 paper.
This section contains tutorials that show you how to use MOPAC via the MOPAC-GUI. As the MOPAC-GUI is very similar to the ADF-GUI, we will focus here on MOPAC specific features.
If you are not familiar with the ADF-GUI (or the MOPAC-GUI), please have a look at the Introduction and ADF-GUI Overview tutorials.
For example, it will show you how to start the MOPAC-GUI.
MOPAC (Molecular Orbital PACkage) is a semiempirical quantum chemistry program based on Dewar and Thiel's NDDO approximation. It is available from OpenMOPAC. The documentation for MOPAC is also available on that site.
MOPAC is included with the ADF distribution. However, it needs to be enabled in your license file. If it is not enabled, please contact SCM to get more information. Note that MOPAC is free for academic users. The MOPAC that is included with the ADF distribution is just the standard MOPAC from OpenMOPAC, and is updated frequently.
If you wish to use a MOPAC version different from the one included with the ADF distribution, you can do this by setting the SCM_MOPAC environment variable, either in your shell startup script or via the SCM → Preferences command:
The MOPAC-GUI is basically the same set of programs as the ADF-GUI. we suggest you start with the ADF-GUI tutorials if you are not familiar with the ADF-GUI.
When saving a job and running, MOPAC will be executed. The MOPAC results (output and AUX file) will be converted to a .rkf result file, that allows visualization of the results with the GUI modules like ADFview, ADFmovie and ADFspectra. Results that may be visualized include geometry, optimization changes (with energy), charges, bond orders, frequencies and normal modes. Obviously the MOPAC output file is also available
This tutorial will give you an example how to use MOPAC via the MOPAC-GUI from SCM.
Start ADFjobs SCM → DFTBinput In the DFTBinput window: Panel bar DFTB → Mopac (thus, click on 'DFTB', and select Mopac from the menu)
We will now make a Toluene molecule. Most conveniently this is done by simply searching for it within the MOPAC-GUI:
Control/Command-F Enter 'toluene' as search text Select 'Toluen' (the second match) from the list of matches Click once in the drawing area in empty space to clear the selection
In the Mopac Main panel you can see what options for MOPAC have been selected: by default a Geometry optimization will be performed, using the PM6 method. MOPAC will also calculate the bond orders, and the GUI will automatically update the molecular bonds using these calculated bond orders. To show this, we will first remove the bonds and then run MOPAC by pressing the Run button:
Bonds → Remove Bonds Press the 'Run' button
The geometry has been optimized, but as the starting geometry was very reasonable this did not lead to big changes. In the status area you can see that MOPAC used 18 iterations, and you can see the final heat of formation. And obviously the bonds reappeared. These bonds are actually based on the bond orders as calculated by MOPAC.
You can now also show the calculated atomic charges:
View → Atom Info → Mopac Charges → Show
To see the MOPAC output, check the 'Show output' check box. And there is another way to run MOPAC interactively: use the pre-optimizer button (the cog wheel).
Check the 'Show output' check box Right-click on the cog-wheel (the pre-optimizer button), and select Mopac from the menu
As the geometry already had been optimized, nothing much happens. Except the MOPAC output file will be opened and appear in a new window. This file actually starts with the AUX file, followed by the output file.
Close all windows except ADFjobs (no need to save your setup)
The previous section showed how to run MOPAC interactively, typically to be used as a pre-optimizer. However, you can also run the MOPAC calculation as a 'real' job. Then the result files will be saved, and you can analyze them with the GUI tools.
SCM → DFTBinput In the DFTBinput window: Panel bar DFTB → Mopac (thus, click on 'DFTB', and select Mopac from the menu) Control/Command-F Enter 'toluene' as search text Select 'Toluen' (the second match) from the list of matches Click once in the drawing area in empty space to clear the selection Check the 'Show output' check box
Once again we have our Toluene molecule. Now to run it as a job simply use the Save and Run command from the menu:
File → Save, save the job with name 'Toluene' File → Run
When asked to read the new coordinates (when the calculation finishes), click Yes SCM → Movie
The logfile will show the progress of the calculation:
SCM → View Properties → Atom Info → Atom Charge: Net (mopac) → Show
As the 'Show Output' check box has been checked, the Mopac output is available:
Close the ADFview window SCM → Output Scroll to the end
Now for a geometry optimization it might be interesting to see a movie of the progress of the optimization:
Close the ADFoutput window SCM → Movie Graph → Energy
You can also use ADFmovie to get detailed information about the geometry (distances, angles and so on), just as for an ADF geometry optimization. If you don't know how, check the ADF-GUI tutorials.
Close all windows except ADFjobs and the ADFinput window showing your MOPAC input
Now we will modify our MOPAC job to perform a frequencies calculation:
Activate the ADFinput window (MOPACinput) In the Mopac Main panel, change the Task to 'Frequencies' File → Run Answer 'Yes' to save your setup Answer 'Yes' to acknowledge that you are saving over a previous job
The logfile will be shown, and will very soon show the job is ready,
SCM → Spectra
The calculated IR spectrum will be shown:
Right-click on the peak around 2762 cm-1, and select mode A_37 (with the maximum intensity)
The selected mode will be shown:
In ADFspectra: tear off the A menu by selecting the dashed line (A → - - - - ) Use the torn-off menu to examine some more modes SCM → Close All
This section contains tutorials that show you how to use ReaxFF via the ReaxFF-GUI. As the ReaxFF-GUI is very similar to the ADF-GUI, we will focus here on ReaxFF specific features.
If you are not familiar with the ADF-GUI (or the ReaxFF-GUI), please have a look at the Introduction and ADF-GUI Overview tutorials.
For example, it will show you how to start the ReaxFF-GUI, and how to manually edit molecules.
The first tutorial provides you with an overview over all relevant GUI modules. It will create a mixture of methane and oxygen, and burn it. This will clearly demonstrate that ReaxFF can handle reactions.
The second tutorial will show you how to set up a more complex system and how to use some of the advanced features of ReaxFF. In setting up the system you will also find out how to create an aluminum surface.
This tutorial will help you to:
Start ADFjobs SCM → ReaxFF input
Note that ReaxFFinput and ADFinput are actually identical. By starting ReaxFFinput from the SCM menu ADFinput will immediately switch to ReaxFF mode.
Next we will make the methane - oxygen mixture. For full combustion we need at least 2 oxygen molecules for each methane molecule. So we will use 100 methane molecules and 250 oxygen molecules.
Edit → Builder
The Builder allows you to build your system, and set some things like the cell vectors that define the computation cell. ReaxFF always uses periodic boundary conditions. The default cell is a cube with sides of 100 Angstrom.
You can use the Builder to add many molecules, randomly distributed. In the list of molecules to be added the Current molecule is already present. If you have any molecules already, the new molecules will be added. Right now we have no molecules yet, so we can remove that entry.
Click the '-' button in front of 'Current' Click the + button in front of 'Molecules' once (so now we can specify two kinds of molecules to add) Change the box to a cube of 25.0 Angstrom
Click in the entry field of the first line Type 'me' (without the quotes) to search for methane
As you can see, a search box appears to find your molecule, very similar to the search box in the panel bar.
Select the 'Methane (ADF)' match Click the file select button on the other 'Fill box with' line Click in the entry field of the second line Type 'o', to search for O2 Select the 'O2 (ADF)' match Change the 100 copied of O2 into 250 copies
You now have specified how the builder should build your system: 100 methane molecules with 250 oxygen molecules added.
At the bottom of the Builder panel you can see that the current density is zero. The new density, after adding methane and oxygen, will be around 1 g/mL, which is obviously very high for this mixture. For this tutorial that is fine as it means things will happen faster.
Next we will actually generate the molecules:
Click the 'Generate Molecules' button
The molecules are generated at random positions and orientations, with constraint that all atoms (between different molecules) are at least the specified distance (2.5 Angstrom) apart.
The system looks good, we can now close the Builder:
Click the Close button at the bottom of the Builder window
The next step is to set up the details of the simulation. For this tutorial we will perform an MD simulation, at very high temperature and density. This is to make things happen quickly. Obviously it is not a realistic system.
Select the 'Molecular Dynamics' task Click the 'i' on the right side of the Force field line
A new window should appear describing what force fields are available, including a short description and references. For this particular example we will use the CHO force field for hydrocarbon oxidation.
Close the window describing the force fields Click on the folder-like button in the Force field line Select the CHO.ff force field Enter 50000 as number of iterations Enter 10000 as number of non-reactive iterations Specify a temperature of 3500 K
Now we will run your set-up:
Use the File → Run command When asked to save your input, save it with the name 'Methane'
ADFjobs should come to the foreground, and your job should be visible at the top. On the right side you can see that the job is running (this is indicated by the gear-icon). When running, also a window will appear showing the progress of your simulation (the logfile):
As you can see in the logfile, the simulation is running.
To see more details, we now will use ADFmovie. Note that you can do this while the simulation is still running!
Start ADFmovie: SCM → Movie in the logfile window
ADFmovie will show you the trajectory of your system. Note that it will automatically read new data as soon as it becomes available.
It can also show graphs of the properties that ReaxFF calculates:
In the ADFmovie window: Use the Properties → Temperature command Use the Properties → Temperature (Set) command Use the Properties → Potential Energy command Make the window bigger to make viewing the graph easier
Note that the two temperatures (the actual temperature, and the set temperature) are both plotted on one axes. The other axes is used for the second property (in this case the potential energy).
You can go to a particular point in the simulation using the slider below the window showing your system, or you can click somewhere on one of the curves plotted. You can also use the arrow keys (left and right) to move through the simulation.
Click on the temperature curve, move around in the movie Jump to the end of the movie to follow the progress
As ReaxFF is a reactive force field, reactions may take place. In this particular example the methane should react with the oxygen, eventually producing H2O and CO2.
You can make graphs that show how many of the different molecule types are present. The following instructions often work, but it depends on what molecules are present in your simulation. You might try this step again after waiting some time. Remember that we requested 10000 non-reactive iterations, so just leave it running for 30000 iterations or so. Especially the production of H2O and CO2 take some time.
After about 30000 iterations: Use the Properties → Molecule Fractions command In the window that pop-ups, select H2O Click Add / Remove Repeat these three steps for CH4, and for CO2 if any present yet.
Obviously, no reactions take place in the first 10000 iterations.
You can put one of the curves on a different axes if you wish:
Click once on the curve showing the number of CH4 molecules, this makes it the 'active' curve Use the Graph → Curve On Right Axes command
Clicking on the curve also had two other effects (besides making it the active curve): you jumped to the iteration in the movie corresponding to the point where you clicked, and the molecules that belonged to that curve are selected.
The selected molecules are shown with Balls and Sticks, instead of wire-frame only, to make it easier to spot them. Flying to the selection also makes it easier to spot them:
Use the Properties → Molecule Fractions command to add a CO2 curve if you have not yet done so Click on the curve showing the CO2 production Use the View → Fly To Selection command a few times
When you now go forward or backwards in time, it is easier to see how the reactions actually take place. Note that the atoms remain selected, even if they are no longer part of a CO2 molecule. In a similar way you can focus on H2O produced:
Click on the curve showing the H2O production Use the View → Fly To Selection command if needed
Wait until the calculation is ready Select the CO2 curve (click on it) Move it to the left axes: Graph → Curve On Left Axes
To prepare for the next tutorial, quit everything:
Bring the ADFjobs window to the foreground Use the SCM → Close command to close all windows for this job
This tutorial will help you to:
Start ADFjobs Use the SCM → ReaxFF input menu command
Bulk aluminum has an fcc crystal structure, with a lattice constant of about 4.0 Angstrom.
To create the surface, we first build bulk aluminum:
Edit → Crystal → Cubic → fcc
In the dialog that appears:
Set the lattice parameter to 4.0
Set the atom type to Al
Click the 'OK' button
View → Periodic → Repeat Unit Cells
In your molecule editor screen you should see a picture of the bulk aluminum structure. The unit cell contains one atom. As ReaxFF input by default does not show repeated cells, it had to be turned on explicitly. Otherwise you see only the single atom in the unit cell.
Now create the surface:
Edit → Crystal → Generate Slab...
In the Dialog window:
Check the 'cartesian' check box
Set the number of layers to 3
Click OK
View → Periodic → Repeat Unit Cells
You will see three atoms as it is a three-layer slab.
In this case we do not want just one unit cell, but a much bigger piece of the slab:
Edit → Crystal → Generate Super Cell... Put 10, 10, 1 on the diagonal Click OK Rotate get a better view
Now we have a real slab of aluminum, three layers thick. Note that the Generate Super Cell command adjusted the lattice vectors to match the new unit cell.
The next step is to add water to the system, and make sure that the slab is in the center of the simulation box:
Edit → Builder
Change the third dimension of the lattice vectors to 100.0 Check the 'Move to center of box' box on the line for the Current molecule Type 'wa' in the line with 'Fill box with' Select 'Water (ADF)' from the search results Specify 2500 copies Specify a distance of 1.0 Angstrom (this is the minimum distance between the molecules) Click the 'Generate Molecules' button on the bottom Use the View → Periodic → Show Unit Cell command to visualize the box Rotate to your favorite view
Now we will set up the MD-simulation. We will use the Al-water force field included in the undocumented directory:
Close the Builder by clicking the Close button on the bottom Select the 'Molecular Dynamics' Task Click on the folder icon to select a force field Select the undocumented/Al-water.ff Specify 100000 iterations
In the simulation we would like something to happen. So we will use a high temperature for water, but try to keep the aluminum cool. Also, we will start with a low temperature MD to relax the initial set-up. This can all be accomplished using a temperature regime, specifying different temperatures for different regions.
For this we first need to define two new regions: one for the aluminum slab, and one for the water. Regions in ReaxFFinput are the same as in the ADF-GUI: they are defined as a collection of atoms. Via the Regions panel you can define the regions:
panel bar Model → Regions Click once on the select button (the V like button) in the Auto-Generated region
By pressing the select button you have selected all atoms in the 'Auto-Generated' region. This region will always contain the atoms that are added by the Builder. Thus, as you can see, all water molecules are selected.
We are going to set up a region for the water molecules, and a region for the aluminum surface:
With the water molecules still selected, press the + button in front of the Regions label Click and select the text 'Region_2' and change it into 'Water'
Now we have a region defined that contains all water molecules. Note that the selected atoms (in this case all water molecules) where automatically added to the new region. The + and - buttons add or remove the selected atoms to a region.
Click in the drawing area so it has focus (otherwise the next will select the region name)
Select → Select All
Press the + button in front of the Regions label again to add a new region
(containing everything as everything was selected)
Change the name Region_3 to Al
Click the select button in the Water region to select all water molecules
Click the '-' button in the Al region to remove the water from the Al region
Click in empty space to clear the selection
The different regions are shown with colors. This does make the display a lot slower. You can turn off (or on) this display with the check box in front of the Region names, and you can set visualization style per region:
Uncheck the check box at the left of the Water region line Press the triangle on the right side of the Al region line, and choose for the 'Balls And Sticks' visualization option
Now we have defined the regions that we need, now set up the temperature regime:
Panel bar Model → Temperature Regime Press the '+' button in front of the Regime label twice to add two temperature steps
At iteration 0, we start with an MD simulation at T=0 K:
change the first 298 to 0 (note this applies to region All which is everything)
At iteration 20000, we set the water temperature to 2000, while keeping the aluminum temperature at 0:
Change the 10000 to 20000 Press the + button on the right side of the line starting with 20000 Change the first All on the 20000 step to Al, and change its temperature to 0 Change the second All on the 20000 step to Water, and change its temperature to 2000
Your temperature regime setup should look like this:
Now we can run our set up:
File → Run When asked to save, specify Al-water as filename Let it run for some time (until iteration 500 or so) SCM → Movie View → Molecule → Balls And Sticks Rotate and zoom to get a good view of the surface
If you look around, you will most likely see many water molecules attaching to the aluminum surface. After some time, you might also see hydrogen atoms distracted from the water molecules, which are now moving around the surface on their own.
You might leave the simulation running to see what will happen. It will take a lot of time though.
If you do not want to wait for the simulation to finish, kill the job:
Bring the ADFjobs window to the front Make sure your Al-water job is selected (click once on it if not) Kill it Job → Kill Close all GUI modules: SCM → Close All
This part will provide a hands-on tutorial demonstrating the COSMO-RS GUI. Its purpose is to give a quick tour of the main features of the COSMO-RS GUI.
The first tutorial describes how to create COSMO result files that are needed as input for the COSMO-RS calculation. The second tutorial describes some analysis tools (sigma profile, sigma potential), and shows how to set COSMO-RS parameters. The third tutorial describes how to calculate properties, such as solvent vapor pressures, solvent boiling points, partition coefficients (log P), activity coefficients, solvation free energies, Henry's law coefficients, solubilities, vapor liquid diagrams binary mixture (VLE/LLE), and excess energies. The fourth tutorial describes how to use the COSMO-RS database ADFCRS-2010 of COSMO result files. The fifth tutorial describes how to calculate pKa values
Additional information may be found in the COSMO-RS part of the GUI Reference manual and COSMO-RS manual
If you already have COSMO result files for all the compounds that you are interested in you can skip this tutorial, without problems of continuity. For example, ADF has a database of COSMO result files, the COSMO-RS database ADFCRS-2010.
The purpose of this tutorial is to teach you how to make data for a compound using the ADF program such that it can be read by COSMO-RS. COSMO-RS expects so called COSMO result files, which are results of quantum mechanical calculation using COSMO. In ADF such a COSMO result file is called a TAPE21 (.t21) file, or a COSKF (.coskf) file. For example the COSMO-RS database ADFCRS-2010 consists of .coskf files. In other programs such a file can be a .cosmo file. For example, at http://www.design.che.vt.edu/VT-Databases.html a database of .cosmo files can be found, which were made with a different program. Note that the optimal COSMO-RS parameters may depend on the program chosen.
Please read through the first ADF-GUI tutorial before starting with this tutorial,
see the ADF-GUI tutorials.
Even better: try using the ADF-GUI yourself, especially
Tutorial 1
In this tutorial an ADF COSMO result file and a MOPAC COSMO result file is made. For ADF COSMO-RS calculations the recommended choice is to use ADF COSMO result files.
For this tutorial we prefer to work in a separate directory, for example a directory called Tutorial, as was explained in ADF-GUI Tutorial 1.
You know how to start ADFjobs (in your home directory), and move to the Tutorial directory:
Start adfjobs Click on the Tutorial folder icon
Next start ADFinput using the SCM menu.
Select the SCM → ADF input menu command.
First we construct a water molecule, and preoptimize its geometry:
Select the O-tool by clicking on the button with the 'O' Click somewhere in the drawing area to create an oxygen atom Select the select-tool by clicking on the button with the arrow Click once in empty space so nothing is selected Select Atoms → Add Hydrogen Click the optimizer button
Your water molecule should look something like this:
The next step is to optimize the geometry using ADF. In this case we will use the Becke-Perdew functional and a TZP basis set, since those are used for the optimization of the COSMO-RS parameters.
Enter a proper title in the Title field (like 'water') Select Preset → Gas Phase CRS
For your information, the proper settings for the gas phase geometry optimization are: The Becke Perdew exchange correlation functional (GGA:BP), use of the scalar relativistic ZORA Hamiltoninan, a TZP small core basis set (for Iodine a TZ2P small core basis set), and an integration accuracy of 6. Like for Iodine for heavier elements than krypton, a TZ2P small core basis set is recommended.
With the proper options selected, now run ADF:
Select File → Run In the file select box, choose a name for your file (for example 'water_gasphase' to make a distinction in the names of the gas phase molecule and the solvated molecule) and click 'Save'
Now ADF will start automatically, and you can follow the calculation using the logfile that is automatically shown.
Wait until the optimization is ready (should take very little time) Click 'Yes' in the pop-up to read the coordinates from a .t21 file.
Now the geometry of the water molecule is the optimized one.
First set up the COSMO calculation with the COSMO parameters as they were used in the ADF parametrization of COSMO-RS. The molecule is in a solvent, represented in the COSMO model by a dielectric medium with a dielectric constant. This dielectric constant is infinite in the COSMO-RS model.
Select Preset → Solvent CRS
Part of the COSMO parameters that are set can be found if one selects Model → Solvation. The dielectric constant is infinite if one chooses 'CRS' for the Solvent. The so called 'Klamt' radii are used for the construction of the molecular cavity.
Another part of the COSMO parameters that are set can be found if one selects Details → COSMO:
For the COSMO-RS calculation it is important that both the energy of the gas phase molecule and the energy of the solvated molecule are known. Also important is to use the same charge and/or spin-polarization in the gas phase molecule as in the solvated molecule. In order to have both values a restart is required with the .t21 file of the gas phase molecule as restart file for the solvated molecule. Note that this is a different procedure than in the ADF2008.01 tutorial.
Use the panel bar Details → Files (Restart) command Click the file select button next to 'Restart file', a popup window will appear Select water_gasphase.t21 (the .t21 file of the gas phase molecule) Click 'Open' in the pop-up
For sake of clarity we will save the COSMO calculation under a different name:
Select the 'Save As..' command from the 'File' menu In the file select box, choose a name for your file (for example 'water') and click 'Save'
With the proper options selected, now run ADF:
Select File → Run
Now ADF will start automatically, and you can follow the calculation using the logfile that is automatically shown.
Wait until the COSMO calculation is ready (should take very little time) Click 'Yes' in the pop-up to read the coordinates from a .t21 file.
After the calculation has finished the file water.coskf, which is an ADF COSMO result file, can be used as input for a COSMO-RS calculation.
Note that a .coskf file is not a complete .t21 file. For example, if one has such a .coskf file, only the COSMO surface charge density can be viewed with ADFview. Thus a .coskf file is mostly useful for COSMO-RS calculations.
A MOPAC COSMO result file can be created in almost the same way as an ADF COSMO result file. In this step we will skip the first steps of creating the water molecule. We will change the program from ADF to MOPAC, and select the COSMO solvation method.
Select ADF → Mopac Select Solvation Method → COSMO-CRS
Part of the COSMO parameters that are set can be found if one selects Details → Run Script. Technically ADFinput will add some keywords to the input for MOPAC (see also http://www.openmopac.net) in case of a MOPAC COSMO calculation:
PM6 EPS=9999.9 RSOLV=1.3 COSWRT NSPA=362
For sake of clarity we will save the COSMO calculation under a different name, and run the calculation
Select the 'Save As..' command from the 'File' menu In the file select box, choose 'water_mopac' as name for your file and click 'Save' Select File → Run Wait until the optimization is ready (should take very little time) Click 'Yes' in the pop-up to read the coordinates from a .rkf file.
After the calculation has finished the file water_mopac.coskf, which is a MOPAC COSMO result file, can be used as input for a COSMO-RS calculation.
Note that MOPAC is a semi-empirical quantum chemistry program, whereas ADF is based on density functional theory (DFT). Thus the MOPAC COSMO result files will not be of the same quality as the ADF COSMO result files.
For this tutorial it is convenient to start with an empty directory, for example, with the name Tutorial.
In tutorial 1 it was shown how to make ADF COSMO result files. In this tutorial we will use some ADF COSMO result files that were made before. These files can be found the directory $ADFHOME/examples/crs/Tutorial2. Copy these COSMO result files (water.coskf, methanol.coskf, ethanol.coskf, and benzene.coskf) to the directory Tutorial. As the name suggests these are COSMO result files of water, methanol, ethanol, and benzene, respectively.
Note that these COSKF (.coskf) files contain only the part of an ordinary TAPE21 (.t21) file which is needed in a COSMO-RS calculation. These COSKF files can only partly be used in ADFview, for example.
On a Unix-like system, enter the following command:
$ADFBIN/adfcrs &
On Windows, one can start ADFcrs by double-clicking on the ADFcrs icon on the Desktop:
double click the ADFcrs icon on the Desktop
On Macintosh, use the ADFLaunch program to start ADFcrs:
double click on the ADFLaunch icon which will start ADFjobs Select SCM → COSMO-RS
Alternatively if you know how to start ADFjobs (in your home directory), and move to the Tutorial directory:
cd $HOME Start adfjobs Click on the Tutorial folder icon Select SCM → COSMO-RS
This ADFcrs window consists of the following main parts:
This is the so called 'Compounds' window, which is the window you will get if you select Compounds → Compounds.
Select Compounds → Add Compound(s)
A file select box will open , that looks like (may look different on different platforms)
or
Select Files of type (or Filter) → COSMO kf file (*.coskf) Select water.coskf Click 'Open'
On the right side of the 'Compounds' window one finds some data that was written from the file water.coskf. Here it is also possible to add some pure compound input data. We will not do so now since there is no need. For some other types of compounds user input is required at this point, however. We will encounter an example of that in the next step.
Repeat these instructions to add methanol.coskf, ethanol.coskf, and benzene.coskf.
In the COSMO-RS model (see Ref.[2] in the COSMO-RS manual) there is a ring correction term. This is important for, for example, the benzene molecule, which has 6 ring atoms. However, it is really required only when the vapor pressure of the compound is going to be computed (either because that is explicitly requested or because it is used in predicting partial vapor pressures in a mixture or gas/liquid partitioning coefficients).
Click on the left side benzene Enter '6' without quotes in the 'Nring' field on the right side
For some properties, like solubility of a solid, one can include some pure compound properties in the left window of the 'Compounds' window for a selected compound.
COSMO-RS
Select Method → COSMO-RS Select Method → Parameters
Default 'ADF combi2005' COSMO-RS parameters are selected, which are ADF optimized COSMO-RS parameters. See also a discussion of the COSMO-RS parameters in the COSMO-RS manual. If one selects the 'Klamt' option for 'Use: ... COSMO-RS parameters', the optimized parameters are chosen, which are optimized by Klamt et al., see Ref. [2] in the COSMO-RS manual.
COSMO-SAC
Select Method → COSMO-SAC Select Method → Parameters
Only the activity coefficients are calculated according to the COSMO-SAC method. Vapor pressure will be approximated using ideas from the COSMO-RS method, and not using the COSMO-SAC method. In order to do that some COSMO-RS parameters can be set that are used in an ADF COSMO-SAC calculation. See also a discussion of the COSMO-SAC parameters in the COSMO-RS manual.
Select Method → COSMO-RS
You can use ADFview to have a look at the COSMO surface, and the COSMO surface charge density. This is possible if the COSMO result file of the compound is a .coskf file or a .t21 file.
Select Compounds → Compounds
Click on the left side water
Select Compounds → Show Selected Compound
Select (in the ADFview window that pop-ups) Add →
COSMO: Surface Charge Density → on COSMO surface (reconstructed)
Then you will see something like:
The red part represents positive COSMO charge density (the underlying molecular charge is negative), the blue part negative COSMO charge density (the underlying molecular charge is positive). You can also look at the COSMO surface points themselves.
Select Add → COSMO: Surface Charge Density → on COSMO surface points
The small spheres represent the COSMO surface points that are used for the construction of the COSMO surface.
Next we will close this ADFview window.
Select the ADFView window 'water' Select File → Quit Select the COSMO-RS GUI window
ADFview has many options to change the look of the picture.
Select Compounds → Compounds
Click on the left side methanol
Select Compounds → Show Selected Compound
Select (in the ADFview window that pop-ups) Add →
COSMO: Surface Charge Density → on COSMO surface (reconstructed)
Select View → Background → White
Check View → Anti-Alias
Increase the size of the ADFView window, such that a control line
at the bottom of the ADFview window is visible
In the control line click on the 'Cosmo surface' pull-down menu and
use the Show Details command
Change the Opacity to 70
Next we will close this ADFview window.
Select the ADFView window 'methanol' Select File → Quit Select the COSMO-RS GUI window
Select Analysis → Sigma Profile In the left part of the window select 'water' for compound '1' Select 'ethanol' for compound '2' Select 'benzene' for compound '3' Select 'methanol' for compound '4' Select File → Run Click 'Yes' when asked to 'Save changes (required to run)?' Enter the name 'tutorial2' in the 'Filename' field
The sigma profiles (σ-profile) of the three pure compounds will be shown in a graph and in a table in the right part of the window. The whole window can be resized. The relative size of the left part of the window compared to the right part can be changed if one moves the sash that is in between these parts. In the right part of the window one can also change the relative size of the upper part compared to the lower part if one moves the sash that is in between these parts.
The σ-profile shows the amount of surface area for a given COSMO charge density. If the default settings are used, the red curve is the water σ-profile, the blue curve the methanol σ-profile, the green curve is the benzene σ-profile, and the yellow curve the methanol σ-profile. If one clicks in the graph one can zoom (right mouse, or Command-Left (drag up or down)), or translate (left mouse) the graph. If one clicks in the graph window at the left or below the axes, a popup window will appear in which one can set details for the graph window.
Note that the σ-profile depends on the method (COSMO-RS or COSMO-SAC) that was used in the calculation. Here we have used COSMO-RS. In this case the σ-profile depends on the actual value for r_av (rav), which is one of the COSMO-RS parameters, see one of the previous steps.
One can also look at the hydrogen bonding part of the σ-profile.
In the left part of the window select 'Pure: HB' for Solvent Press 'Run'
Select Analysis → Sigma Potential In the left part of the window select 'water' for compound '1' Select 'ethanol' for compound '2' Select 'benzene' for compound '3' Select 'methanol' for compound '4' Press 'Run'
The sigma potentials (σ-potential) of the three pure compounds will be shown in a graph and in a table in the right part of the window. The σ-potential depends on the temperature of a compound. Here the temperature is set to 25 °C (298.15 K).
If the default settings are used, the red curve is the water σ-potential, the blue curve the methanol σ-potential, the green curve is the benzene σ-potential, and the yellow curve is the methanol σ-potential. In the details for the graph window the line widths for all curves were set to '3'.
Note that the σ-potential is not calculated for values of the COSMO charge density that are non-existent on the COSMO surface of a certain compound.
How the properties are calculated and definitions used can be found in the section Calculation of properties in the COSMO-RS manual.
For this tutorial we assume that you know how start ADFcrs and how to add compounds. Like in tutorial 2 we ask you to add the compounds water, methanol, ethanol, and benzene. One can do this, for example, by opening the .crs file that was created in tutorial 2. Save the file as tutorial3.crs.
Start ADFcrs Select File → Open... Select 'tutorial2.crs' in the 'Filename' field Select File → Save As Enter the name 'tutorial3' in the 'Filename' field
Alternatively on a Unix like system one may copy the COSMO result files (water.coskf, methanol.coskf, ethanol.coskf, and benzene.coskf) in the directory $ADFHOME/examples/crs/Tutorial2 to an empty directory and enter the following command in this directory where the COSMO result files are present:
$ADFBIN/adfcrs water.coskf methanol.coskf ethanol.coskf benzene.coskf &
Note that one has to set the number of ring atoms for the benzene compound.
Select Compounds → Compounds Click on the left side benzene Enter '6' without quotes in the 'Nring' field Select File → Save As Enter the name 'tutorial3' in the 'Filename' field
In the compounds window one can also set the vapor pressure of the pure compounds at a given temperature, or set the Antoine parameters. If these values are not specified (if they are zero) then the pure compound vapor pressure will be approximated using the COSMO-RS method. This is relevant, for example, for the calculation of the (partial) vapor pressures of mixtures, calculation of boiling points of mixtures, and calculation of Henry's law constants.
The vapor pressure of a solvent at different temperatures can be calculated with Properties → Solvent Vapor Pressure.
Select Properties → Solvent Vapor Pressure Select 'methanol' for the first component in the Solvent Press 'Run'
In this case the result is a table with one entry:
Select Celsius for the unit of temperature Enter '0.0' without quotes in the 'Temperature from:' field Enter '100.0' in the 'to:' field Press 'Run'
In this case the result is a graph and a table. If one clicks in the graph window at the left or below the axes, a popup window will appear in which one can set details for the graph window. After some manipulation in the details of the graph, the graph can look like:
In this case COSMO-RS predicts a vapor pressure of about 0.61 bar (around 455 Torr) at 323.15 K (50.0 °C) for the pure liquid methanol.
The boiling point of a solvent at different pressures can be calculated with Properties → Solvent Boiling Point.
Select Properties → Solvent Boiling Point Select 'methanol' for the first component in the Solvent Select 'ethanol' for the second component in the Solvent Enter '0.5' for the 'Mole fraction' of methanol Enter '0.5' for the 'Mole fraction' of ethanol Select atm for the unit of pressure Enter '0.1' in the 'Pressure from:' field Enter '1.0' in the 'to:' field Press 'Run'
In this case the result (may take several seconds) is a graph and a table.
The red curve is the total vapor pressure, the blue curve the partial methanol vapor pressure, and the green curve is the partial ethanol vapor pressure. The table gives the numerical values.
Thus in this case COSMO-RS predicts a boiling point of 339.8 K (66.7 °C) at 1 atm. for this mixture of 50% mole fraction methanol and 50% mole fraction ethanol. At this temperature COSMO-RS predicts that the vapor consists about 69% of methanol.
Using View → Graph Y Axes → one can view different properties in the graph, like activity coefficients and excess energies.
The flash point of a mixture can be calculated with Properties → Solvent Flash Point, if pure compound flash points are given as input.
Here we mix equal volumes of water (assuming a density of 0.997 kg/L) and ethanol (assuming a density 0.789 kg/L). For a flash point calculation the pure compound flash points are needed as input, since COSMO-RS does not predict pure compound flash points. The ADF COSMO-RS module uses Le Chatelier's mixing rule to calculate the flash point of a mixture.
Select Compounds → Compounds Select 'ethanol' Enter '286' for the 'Flash point' of ethanol Select Properties → Solvent Flash Point Change the popup menu 'Mole fraction' in 'Mass fraction' Select 'ethanol' for the first component in the Solvent Select 'water' for the second component in the Solvent Enter '0.442' for the 'Mass fraction' of ethanol Enter '0.558' for the 'Mass fraction' of water Press 'Run'
In this case the calculated flash point will be close to 25 °C.
The activity coefficients of infinitely diluted solutes in a solvent can be calculated with Properties → Activity coefficients.
Select Properties → Activity coefficients Select 'water' for the first component in Solvent Press 'Run'
If one does not supply a density of the solvent in the input the program calculates the density of the solvent by dividing the mass of a molecule with its COSMO volume. Note that the calculated activity coefficients do not depend on this density. One may improve the results for the calculation of the Henry constants, if one includes pure compound vapor pressures in the 'Compounds' window (Compounds → Compounds).
The result of the calculation is given in the form of a table.
Relevant for the calculation of ΔG from the gas phase to the solvated phase is the reference state, used here is 1 mol/L both in both phases.
The partition coefficients (log P) of infinitely diluted solutes in a mixture of two immiscible solvents can be calculated with Properties → Solvent Log Partition Coefficients.
Select Properties → Solvent Log Partition Coefficients Select 'benzene' for the first component in the solvent Select 'water' for the second component in the solvent Click the check box 'Use input volume solvent phase 1/phase 2' Enter '4.93' in the 'Use input volume solvent phase 1/phase 2' field Press 'Run'
In case of partly miscible liquids, like, for example, the octanol-rich phase of octanol and water, both components have nonzero mole fractions.
Here an input value is used for the volume quotient of the two solvents. If one does not include such value, the program will use the COSMO volumes to calculate the volume quotient. The COSMO volumes can be found in the 'Compounds' window.
The result of the calculation is given in the form of a table.
The solubility of a solute in a solvent can be calculated with Properties → Solubility. The solute can either be a liquid, solid, or gas.
First the solubility of benzene in water for a range of temperatures.
Select Properties → Solubility Select 'water' for the first component in 'Solvent' Select 'benzene' for the first 'Solute' Enter '273.15' without quotes in the 'Temperature from:' field Enter '373.15' in the 'to:' field Press 'Run'
If a range of temperatures is requested a graph is shown.
The assumption made in the solubility calculation may be invalid in case of a solubility of a liquid in a solvent, especially if the solubility of the solvent in the liquid is high. For binary systems one may check this by calculating the miscibility gap in the binary mixture of the two liquids. An example is given for the calculation of the miscibility gap in the binary mixture of Methanol and Hexane.
Note that experimentally benzene is a solid below 5.5 °C, and a gas above 80.1 °C. This has not been taken into account yet.
For the solubility of a solid compound it is necessary to include the melting point, the enthalpy of fusion, and optionally, since it is often not so important, the Δ heat capacity of fusion of the pure compound. These values can be given for each compound when 'Compounds' is selected from the 'Compounds' menu. Here, some experimental values will be included for benzene (see, for example, http://en.wikipedia.org/wiki/Benzene).
Select Compounds → Compounds Click on the left side benzene Enter '278.7' in the 'Melting point' field Enter '2.37' in the 'Δ_fusion H' field
Also an experimental value for the density of water will be used:
Select Properties → Solubility Select 'water' for the first component in 'Solvent' Select 'benzene' for the first 'Solute' Click the check box 'Use input density solvent (kg/L)' Enter '1.0' in the 'Use input density solvent (kg/L)' field Change one of the popup menu's 'Kelvin' in 'Celsius' Enter '0.0' without quotes in the 'Temperature from:' field Enter '10.0' in the 'to:' field Select View → Graph Y Axes → solubility (g/L solvent) Press 'Run'
A graph (and table) is shown, which after some manipulations could look like:
For the solubility of a gas one should click the check box 'Solubility gas' and enter a partial pressure in the 'Pressure' field.
Select Compounds → Compounds Click on the left side benzene Enter '1.01325' in the 'Pure compound vapor pressure' field Enter '353.3' in the 'at temperature' field Select Properties → Solubility Select 'water' for the first component in 'Solvent' Select 'benzene' for the first 'Solute' Click the check box 'Use input density solvent (kg/L)' Enter '1.0' in the 'Use input density solvent (kg/L)' field Use 'Kelvin' as unit for the temperature Enter '353.3' without quotes in the 'Temperature from:' field Enter '373.15' in the 'to:' field Click the check box 'Solubility gas' Use 'atm' units for the pressure Enter '1.0' for the partial vapor pressure of benzene Select View → Graph Y Axes → solubility (mol/L solvent) Press 'Run'
A graph (and table) is shown, which after some manipulations could look like:
The solubility of a gas in a solvent can also be calculated using Henry's law, which is valid for ideal dilute solutions,
Select Properties → Activity coefficients Select 'water' for the first component in Solvent Click the check box 'Use input density solvent (kg/L)' Enter '1.0' in the 'Use input density solvent (kg/L)' field Use 'Celsius' as unit for the temperature Enter '90.0' for 'Temperature' Select only 'benzene' in the list of 'Solutes (infinite dilute)' Press 'Run'
The calculated Henry constant for benzene (infinite dilute) in water will be close to 0.049 mol/(L atm) at 90 °C.
A phase diagram of a mixture of two components can be calculated with Properties → Binary Mixture VLE/LLE. The binary mixture will be calculated for a list of molar fractions between zero and one. This can be done at constant temperature (isothermal) or at constant vapor pressure (isobaric).
A binary mixture is calculated in which the pure compound vapor pressures are approximated using the COSMO-RS method.
Select Properties → Binary Mixture VLE/LLE Select 'water' for the first compound Select 'methanol' for the second compound Use 'Kelvin' as unit for the temperature Enter '298.14' in the 'Temperature' field Press 'Run'
An activity coefficient plot for water(1) and methanol(2) will be shown.
The results of the calculation are also given in the form of a table, which shows the molar (and mass) fraction of each compound in the liquid, the activity coefficients, the activities, the temperature, the total and partial vapor pressures, the molar fraction of each compound in the vapor (Y), the excess Gibbs free energy GE, the excess enthalpy HE (calculated with the Gibbs-Helmholtz equation), the excess entropy of mixing -TSE, the Gibbs free energy of mixing Gmix, the enthalpy of vaporization ΔvapH (calculated with the Clausius-Clapeyron equation).
These values can also be shown in a graph. The property for the x- and y-axes can be selected from the 'View' Menu. For example, a graph of the excess energies can be shown by:
Select View → Graph Y Axes → excess energies
A plot of the excess energies will be shown.
The red curve is the excess Gibbs free energy GE, the blue curve is the excess enthalpy HE, and the green curve is -T times the excess entropy SE.
A binary mixture is calculated with input data for the pure compound vapor pressures. These can be, for example, experimentally observed pure compound vapor pressures. Note that the calculated partial and total vapor pressures will now depend on these input pure compound vapor pressures.
Select Compounds → Compounds Click on the left side water Enter '0.123416' in the 'Pure compound vapor pressure:' field Enter '322.45' in the 'at temperature:' field
Click on the left side ethanol Enter '0.294896' in the 'Pure compound vapor pressure:' field Enter '322.45' in the 'at temperature:' field Select Properties → Binary Mixture VLE/LLE Select 'water' for the first compound Select 'ethanol' for the second compound Enter '322.45' in the 'Temperature' field Select View → Graph X Axes → x2: molar fraction 2 Select View → Graph Y Axes → partial and total vapor pressures Press 'Run'
A vapor liquid equilibrium (VLE) diagram for water(1) and ethanol(2) will be shown.
The red curve is the total vapor pressure, the blue curve is the partial water vapor pressure, and the green curve is the partial ethanol vapor pressure. One can also change the x and y axes, for example:
Select View → Graph X Axes → x2, y2 Select View → Graph Y Axes → total vapor pressure
Water and benzene do not mix well, thus there will be a miscibility gap.
Select Properties → Binary Mixture VLE/LLE Select 'water' for the first compound Select 'benzene' for the second compound Use 'Celsius' as units for the temperature Enter '50.0' in the 'Temperature' field Enter '100' for 'Number of mixtures'
In this case a a liquid-liquid equilibrium (LLE) is calculated. The number of mixtures for which the binary mixture is calculated should be not too small, otherwise the properties of the 2 immiscible liquids phases will not be so accurate.
If the check box 'force 1 liquid phase within possible miscibility gap' is deselected, then results will be shown in the graph and table only for those compositions of the mixture, which are outside of the miscibility gap. If the check box 'force 1 liquid phase within possible miscibility gap' is selected, then results will be shown also within the miscibility gap, with the unphysical conditions that the two liquids are forced to mix.
A binary mixture is calculated in which the pure compound vapor pressures are approximated using the COSMO-RS method if the input values for the pure compound vapor pressures are zero. Alternative one can click a check box in the 'Method' Menu.
Select Method → Parameters Click on the check box 'use input (Compounds Menu) pure compound vapor pressures(s)' to deselect it
Select Properties → Binary Mixture VLE/LLE Select 'methanol' for the first compound Select 'ethanol' for the second compound Enter '10' for 'Number of mixtures' Select Isotherm, isobar, flash point → isobar Select View → Graph X Axes → x1: molar fraction 1 Select View → Graph Y Axes → Temperature Press 'Run'
The calculated boiling points (may take several seconds) for a binary mixture of methanol(1) and ethanol(2) will be shown.
If one clicks in the graph window at the left or below the axes, a popup window will appear in which one can set details for the graph window. In the graph above the 'Units' for the 'Left Y Axes' were chosen to be 'Celsius'.
Next the same binary mixture is calculated in which experimental pure compound vapor pressures are used.
Select Method → Parameters Click on the check box 'use input (Compounds Menu) pure compound vapor pressures(s)' to select it
Select Compounds → Compounds Click on the left side methanol Enter '1.01325' in the 'Pure compound vapor pressure:' field Enter '338' in the 'at temperature:' field Click on the left side ethanol Enter '1.01325' in the 'Pure compound vapor pressure:' field Enter '351' in the 'at temperature:' field Select Properties → Binary Mixture VLE/LLE Press 'Run'
The calculated graph will look not very different, but it will be more accurate.
A phase diagram of a mixture of three components can be calculated with Properties → Ternary Mixture VLE/LLE. The ternary mixture will be calculated for a list of molar (or mass) fractions between zero and one, for each of the compounds. This can be done at constant temperature (isothermal) or at constant vapor pressure (isobaric).
In this step we will use experimental boiling points as input.
Select Compounds → Compounds Click on the left side water Enter '1.01325' in the 'Pure compound vapor pressure:' field Enter '373.15' in the 'at temperature:' field Click on the left side methanol Enter '1.01325' in the 'Pure compound vapor pressure:' field Enter '338' in the 'at temperature:' field Click on the left side benzene Enter '1.01325' in the 'Pure compound vapor pressure' field Enter '353.3' in the 'at temperature' field Click on the left side ethanol Enter '1.01325' in the 'Pure compound vapor pressure:' field Enter '351' in the 'at temperature:' field
Select Properties → Ternary Mixture VLE/LLE Select 'methanol' for the first compound Select 'ethanol' for the second compound Select 'benzene' for the third compound Use 'Mole fraction' Enter '10' for 'Number of mixtures' Select View → Graph X Axes → x1: molar fraction 1 Select View → Graph Y Axes → x2: molar fraction 2 Select View → Graph Z Colormap → total vapor pressures Select Isotherm, isobar, flash point → isotherm Use 'Celsius' as units for the temperature Enter '70.0' in the 'Temperature' field Press 'Run'
The result will be a graph and a table. In the table one can find the results of the calculation at 55 (=(n+1)(n+2)/2, with n=10) different compositions. At those compositions the table shows the molar (and mass) fraction of each compound in the liquid, the activity coefficients, the activities, the temperature, the total and partial vapor pressures, the molar fraction of each compound in the vapor (Y), the excess Gibbs free energy GE, the excess enthalpy HE (calculated with the Gibbs-Helmholtz equation), the excess entropy of mixing -TSE, the Gibbs free energy of mixing Gmix, the enthalpy of vaporization ΔvapH (calculated with the Clausius-Clapeyron equation).
These quantities can also be shown in the graph as a colormap, in which the color represents the value of the quantity at a certain composition. On the X axes of the graph one can choose the molar (or mass fraction) of one of the compoumds, on the Y axes one can choos the molar (or mass fraction) of another compoumd. The molar (or mass) fraction of the third compound is then fixed, since the sum of the fractions is 1.
In this case the colormap shows the total vapor pressure:
One can improve the quality of the graph by increasing the number of compositions. Note that the number of different compositions for n=20 is 231 (=(n+1)(n+2)/2).
Enter '20' for 'Number of mixtures' Press 'Run'
If one clicks in the graph window at the left or below the axes, a popup window 'Graph details' will appear in which one can set details for the graph window. If one chooses in the 'Z Colormap' part of this popup window as the minimum color magenta, as maximum color red, use 100 as number of colors, and change the minimum and maximum values, then the graph could look like this:
Select Properties → Ternary Mixture VLE/LLE Select 'water' for the first compound Select 'ethanol' for the second compound Select 'benzene' for the third compound Use 'Mole fraction' Enter '20' for 'Number of mixtures' Select Isotherm, isobar, flash point → isobar Use 'bar' as units for the pressure Enter '1.01325' in the 'Pressure' field Select View → Graph X Axes → x1: molar fraction 1 Select View → Graph Y Axes → x2: molar fraction 2 Select View → Graph Z Colormap → total vapor pressures Press 'Run'
The result will be a graph and a table. Note that this may take some time, since isobaric calculations are more expensive than isothermal calculations. Click in the graph window at the left or below the axes. If one chooses in the 'Z Zolormap' part of the 'Graph details' as the minimum color blue, as maximum color red, use 5 as number of colors, change the unit to Celsius, and change the minimum and maximum values, then the graph could look like this:
In addition to the colormap of the temperature, an approximate miscibility gap of the ternary mixture is shown in the graph. In this case, within the miscibility gap there are two immiscible phases of the liquid in equilibrium. The composition of the two phases, which are in equilibrium, can be found at the end points of the tie line that are drawn. The calculated temperatures within the miscibility gap are calculated with the unphysical condition that the three liquids are forced to mix, thus these calculated temperatures (and other quantities) within the miscibility gap should not be used. By inspection of the graph, one can observe that the calculated minimum boiling point (azeotrope) is around 68 °C.
A phase diagram of a mixture of two solvents, which both could be mixtures, can be calculated with Properties → Solvents s1 - s2 Composition Line. The mixture will be calculated for a list of molar (or mass) fractions of the solvents between zero and one, and the compositions of solvent 1 and solvent 2 are linearly interpolated. This can be done at constant temperature (isothermal) or at constant vapor pressure (isobaric).
In this step we will try to investigate one of the tie lines of the ternary mixture of water, ethanol, and benzene in more detail. An attempt is made to use the tie line on which ends the calculated minimum boiling point is found, see the tie line which is below the black line in the next picture:
The compositions of solvents s1 and s2 are chosen where the black line in the picture above crosses the boundary of possible compositions. This means that solvent s1 and solvent s2 are mixtures of 2 compounds. Again experimental boiling points are used in the calculation.
Select Properties → Solvents s1 - s2 Composition Line Use 'Mole fraction' Select 'water' for the first compound Select 'ethanol' for the second compound Select 'benzene' for the third compound Enter '0.0' for the mole fraction of compound 1 of solvent s1 Enter '0.9' for the mole fraction of compound 1 of solvent s2 Enter '0.3' for the mole fraction of compound 2 of solvent s1 Enter '0.1' for the mole fraction of compound 2 of solvent s2 Enter '0.7' for the mole fraction of compound 3 of solvent s1 Enter '0.0' for the mole fraction of compound 3 of solvent s2 Enter '100' for 'Number of mixtures' Select Isotherm, isobar, flash point → isobar Use 'bar' as units for the pressure Enter '1.01325' in the 'Pressure' field Select View → Graph X Axes → s1_x: molar fraction s1 Select View → Graph Y Axes → pure compound activities Press 'Run'
The result will be a table and a graph.
The activities of the pure compounds should be equal at the end point of a tie line a1 = a1',a2 = a2', and a3 = a3'. If we look at the graph with close inspection this is approximately true for the molar fraction of solvent s1 with (approximately) s1_x = 0.007 and s1_x' = 0.91. At a molar fraction of 0.91 of solvent s1 the calculated temperature is approximately 67.9 °C.
Next we will use this temperature of 67.9 °C and look at the Gibbs free energy of mixing. This will also give information about the miscibility gap.
Select Isotherm, isobar, flash point → isotherm Use 'Celsius' as units for the temperature Enter '67.9' in the 'Temperature' field Select View → Graph X Axes → s1_x: molar fraction s1 Select View → Graph Y Axes → Gibbs energy of mixing wrt pure compounds Press 'Run'
The black line was added to show the miscibility gap more clearly. Indeed at 67.9 °C for molar fractions bewteen s1_x = 0.007 and s1_x' = 0.91, the Gibbs free energy of mixing will be lower if the liquid has 2 phases.
Note, that one should use isothermal conditions, if one wants to use the calculated Gibbs free energy of mixing to determine whether there is a miscibility gap. Note also, that no miscibility gap is calculated if one uses Properties → Solvents s1 - s2 Composition Line, even if there is one, like in this case. This is because with the calculated values for only 1 composition line between 2 solvents, that involve more than 2 compounds, in general one does not have enough information to determine the exact miscibility gap.
ADF has a database of COSMO result files, the COSMO-RS database ADFCRS-2010. Assumed here is that the instruction on this page have been followed to download the database.
For this tutorial we assume that you know how start ADFcrs. Next we will add the compounds in the database.
Select Compounds → Add Compound(s) Select Directory → the directory of the downloaded COSMO-RS database ADFCRS-2010 Select Files of type (or Filter) → List of compounds (*.compoundlist) Select all.compoundlist Click 'Open'
After reading some information about all compounds (may take a minute) the GUI window should look like:
For the compounds in this database the number of ring atoms (Nring) is provided, which is needed in the COSMO-RS model (see Ref.[2] in the COSMO-RS manual) in order to calculate a ring correction term.
In this 'Compounds' window it is also possible to add some pure compound input data. Pure compound data could be found, for example, at Refs. [411-413].
The file 'all.compoundlist' is a plain text file, in which on each line a filename of a ADF COSMO result file is given. One could make a different .compoundlist file which contains a only a selection of the compounds.
The compounds could also have been added by adding multiple .coskf files, by selecting 'COSMO kf file (*.coskf)' for the popup menu 'Files of type' (or 'Filter'), and selecting all .coskf files. On Windows it may, however, fail to add more than a few hundred .coskf files at a time in this way. If that fails, one should use the .compoundlist file.
References
[411] Wikipedia
[412] NIST Chemistry WebBook
[413] CHERIC. Korea Thermophysical Properties Data Bank: Pure Component Properties
There are several ways to select a compound in the list. Here several possibilities are discussed. First browsing:
Use the scrollbar to quickly scroll
Click on a compound with the mouse, and move the mouse up or down,
while pressing the mouse
Click on a compound with the mouse, and press the Up or Down key
A separate search window will popup if one clicks the search button. In the 'Compounds' window this search button looks like:
In other windows it will look like:
Select Compounds → Compounds Click on the search button
Click on 'CAS' to sort by CAS registry number Click again to reverse the order Click on 'Nr' to sort by number Click on 'Name' to sort by name Click on 'Formula' to sort by formula Enter 'pentanol' in the 'Search' field Click on 'Cyclopentanol' Click 'OK'
You can use ADFview to have a look at the COSMO surface, and the COSMO surface charge density. Here we use the selected cyclopentanol of the previous step.
Select Compounds → Compounds
Click on the left side 'Cyclopentanol'
Select Compounds → Show Selected Compound
Select Add → COSMO: Surface Charge Density →
on COSMO surface (reconstructed)
Select Add → COSMO: Surface Charge Density →
on COSMO surface points
Next we will close this ADFview window:
Select the ADFView window 'Cyclopentanol' Select File → Quit Select the COSMO-RS GUI window
In this example the Octanol-Water partition coefficients for 7 different compounds are calculated and compared to experiment.
In case of partly miscible liquids, like the Octanol-rich phase of Octanol and Water, both components have nonzero mole fractions. Next an input value is used for the volume quotient of the two solvents. The values for the nonzero mole fractions in the Octanol-rich phase and the volume quotient of the two solvents were taken from Ref. [421]. If one does not include such value, the program will use the COSMO volumes to calculate the volume quotient. The COSMO volumes can be found in the 'Compound' window.
Select Properties → Log Partition Coefficients
Select '1-Octanol' for the first component in the solvent
Select 'Water' for the second component in the solvent
Enter '0.725' for the Mole fraction of 1-Octanol in phase 1
Enter '0.275' for the Mole fraction of Water in phase 1
Enter '1.0' for the Mole fraction of Water in phase 2
Click the check box 'Use input volume solvent phase 1/phase 2'
Enter '6.766' in the 'Use input volume solvent phase 1/phase 2' field
Enter '298.15' for the 'Temperature' in 'Kelvin'
Click the 'None' button below 'Solutes (infinite dilute)'
Click the search button below 'Solutes (infinite dilute)' to add 'Acetamide',
'Methanol', 'Formic acid', 'Diethyl ether', p-Dichlorobenzene,
'Hexamethylbenzene', and '2,2',4,4',5-Pentachlorobiphenyl'
Press 'Run'
Enter the name 'tutorial4' in the 'Filename' field
The result of the calculation (may take a few seconds) is given in the form of a table. Note that not all compounds were selected as solutes, otherwise the calculation would take much more time.
In the database the name 'Ethoxyethane' is used for 'Diethyl ether', and the name '1,4-Dichlorobenzene' is used for 'p-Dichlorobenzene'. In the search window that popups if the search button is clicked, some alternative names are given for some of the compounds.
Next the calculated Octanol-Water partition coefficients are compared to experiment in the following table.
| Solute | experimental log POW[422] | calculated log POW | |
| 1 | Acetamide | -1.16 | -1.20 |
| 2 | Methanol | -0.82 | -0.53 |
| 3 | Formic acid | -0.41 | -0.26 |
| 4 | Diethyl ether | 0.83 | 1.82 |
| 5 | p-Dichlorobenzene | 3.37 | 3.20 |
| 6 | Hexamethylbenzene | 4.61 | 4.96 |
| 7 | 2,2',4,4',5-Pentachlorobiphenyl | 6.41 | 6.47 |
References
[421] S.-T. Lin and S.I. Sandler, Prediction of Octanol-Water Partition Coefficients Using a Group Contribution Solvation Model, Industrial & Engineeering Chemistry research 38 (1999), 4081
[422] Experimental numbers taken from Wikipedia at July 15, 2010, http://en.wikipedia.org/wiki/Partition_coefficient. The experimental numbers may have been determined at a different temperature than 25 °C.
In this example Henry's law constants for 17 different compounds are calculated and compared to experiment.
The Henry's law constants depend on the density of the solvent, Water in this case. If one does not supply a density of the solvent in the input the program calculates the density of the solvent by dividing the mass of a molecule with its COSMO volume. The density of Water at 20 °C (293.15 K) is approximately 0.998 kg/L. Note that the calculated activity coefficients do not depend on the density of Water.
Remark: To only add the compounds needed in this example, one can also open a new COSMO-RS GUI window (SCM → COSMO-RS), copy the $ADFHOME/examples/crs/Tutorial4/tutorial4.3.compoundlist to the directory where the COSMO-RS database is downloaded, and select this file with Compounds → Add Compound(s). This is recommended. The tutorial4.3.compoundlist is a file with a list of compounds that is limited to the coumpounds needed in this example.
Select Properties → Activity coefficients Select 'Water' for the first component in Solvent Enter '293.15' for the 'Temperature' in 'Kelvin' Click the check box 'Use input density solvent (kg/L)' Enter '0.998' in the 'Use input density solvent (kg/L)' field Click the 'None' button next to 'Compounds' Click the search button next to 'Compounds' to add some compounds: 'Acetaldehyde', 'Acetone', 'Acetonitrile', 'Benzene', 'Chloromethane', 'Cyclopentane', 'Dimethyl sulfide', 'Ethanol', 'Formaldehyde', 'Methanol', 'Methyl acetate', 'Methyl bromide', 'Methyl fluoride', 'Methyl iodide', 'Pyridine', 'Thiophene', and 'Toluene' Press 'Run'
The result of the calculation (may take a few seconds, depending on the number of compounds selected) is given in the form of a table.
The Henry's law constants also depend on the vapor pressure of the pure compounds in the gas phase. In the compounds window one can also set these vapor pressures of the pure compounds at a given temperature, or set the Antoine parameters. If these values are not specified (if they are zero) then the pure compound vapor pressure will be approximated using the COSMO-RS method. Best is to include the experimental vapor pressure for a pure compound at the used temperature, thus in this case at 293.15 K.
Select Compounds → Compounds Click on the left side 'Acetaldehyde' Enter '0.968' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Acetone' Enter '0.246' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Acetonitrile' Enter '0.095' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Benzene' Enter '0.100' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Chloromethane' Enter '4.94' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Cyclopentane' Enter '0.346' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Dimethyl sulfide' Enter '0.530' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Ethanol' Enter '0.059' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Formaldehyde' Enter '4.47' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Methanol' Enter '0.129' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Methyl acetate' Enter '0.230' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Methyl bromide' Enter '1.83' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Methyl fluoride' Enter '33.7' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Methyl iodide' Enter '0.443' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Pyridine' Enter '0.021' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Thiophene' Enter '0.082' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Toluene' Enter '0.029' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Click on the left side 'Water' Enter '0.02536' in the 'Pure compound vapor pressure:' field Enter '293.15' in the 'at temperature:' field Select Properties → Activity coefficients Press 'Run'
For some of the compounds the Henry's law constants differ quite substantially.
Experimental determined Henry's law constants might, for example, be found at http://www.henrys-law.org, where a 'compilation of Henry's Law constants for inorganic and organic species of potential importance in environmental chemistry' were listed by R.Sander, and where also an explanation can be found of the many different definitions and units for Henry's law constants.
The calculated Henry's law constants will be compared to experimental values in the next graph and table. The experimental numbers were taken from Ref. [431], where the experimentally determined dimensionless Henry's law constant Hcc is the inverse of the dimensionless Henry's law constant kHcc, that is used in the COSMO-RS module.
Hcc = kH, invcc = 1/(kHcc)
| Solute | experimental Hcc@20°C [431] | calculated Hcc@20°C | |
| 1 | Methyl_bromide | 2.01 10-1 | 2.05 10-1 |
| 2 | Chloromethane | 3.05 10-1 | 3.80 10-1 |
| 3 | Methyl_fluoride | 6.04 10-1 | 9.86 10-1 |
| 4 | Methyl_iodide | 1.70 10-1 | 1.04 10-1 |
| 5 | Cyclopentane | 5.25 100 | 3.27 100 |
| 6 | Benzene | 1.91 10-1 | 1.65 10-1 |
| 7 | Toluene | 2.09 10-1 | 1.89 10-1 |
| 8 | Methanol | 1.37 10-4 | 2.43 10-4 |
| 9 | Ethanol | 1.48 10-4 | 3.70 10-4 |
| 10 | Formaldehyde | 8.61 10-6 | 2.22 10-2 |
| 11 | Acetaldehyde | 2.21 10-3 | 3.72 10-3 |
| 12 | Acetrone | 1.10 10-3 | 0.78 10-3 |
| 13 | Methyl_acetate | 4.02 10-3 | 4.89 10-3 |
| 14 | Acetonitrile | 6.35 10-4 | 6.35 10-4 |
| 15 | Pyridine | 1.14 10-2 | 2.03 10-4 |
| 16 | Dimethyl_sulfide | 6.35 10-2 | 1.82 10-1 |
| 17 | Thiophene | 7.46 10-2 | 6.23 10-2 |
In most cases the calculated Henry's law constants are quite close to the experimental ones, except for Formaldehyde and Pyridine.
The Henry's law constant of Formaldehyde is more than a factor of 103 wrong. The origin of this error is that in Water solution the hydration of Formaldehyde leads to Methanediol, and Methanediol is even the dominant form if one dissolves Formaldehyde in Water. This is not taken into account in the calculation. In Ref. [432] a distinction is made between the apparent and intrinsic Henry's law constants, which differ from each other by approximately a factor of 103 for Formaldehyde, and a factor of approximately 2.4 for Acetaldehyde. In Ref. [432] the intrinsic Henry's law constant for Formaldehyde was determined to be 2.5 mol/(L atm) at 25 °C, which is close to the value of 1.9 mol/(L atm) which was calculated with COSMO-RS, although at a different temperature of 20 °C.
The calculated Henry's law constant of Pyridine is approximately a factor of 102 different than the experimental value in Ref. [431]. However, the experimental values for Pyridine taken from Ref. [433] are 1.1 102 mol/(L atm) and 9.0 101 mol/(L atm), which are not very different from the calculated value with COSMO-RS of 2.1 102 mol/(L atm). Also the experimental values for Pyridine reported in Ref. [434] are in much better agreement with the value calculated with COSMO-RS.
References
[431] J. Staudinger and P.V. Roberts, A critical compilation of Henry's law constant temperature dependence relations for organic compounds in dilute aqueous solutions, Chemosphere 44 (2001), 561
[432] E.A. Betterton and M.R. Hoffmann, Henry's law constants of some environmentally important aldehydes, Environmental Science & Technology 22 (1988), 1415
[433] R. Sander (1999), Compilation of Henry's Law Constants for Inorganic and Organic Species of Potential Importance in Environmental Chemistry (Version 3), http://www.henrys-law.org
[434] M. Bernauera and V. Dohnal, Temperature dependences of limiting activity coefficients and Henry's law constants for N-methylpyrrolidone, pyridine, and piperidine in water, Fluid Phase Equilibria 282 (2009), 100
In this example the solubility of Vanillin in 9 different solvents is calculated and compared to experiment.
Vanillin (4-Hydroxy-3-methoxybenzaldehyde) is a solid at room temperature. For the solubility of a solid, one should include the melting point and the enthalpy of fusion of the solid in the right window of the 'Compounds' window for a selected compound. If known the Δ heat capacity of fusion should also be included, although it is often less important.
Remark: To only add the compounds needed in this example, one can also open a new COSMO-RS GUI window (SCM → COSMO-RS), copy the $ADFHOME/examples/crs/Tutorial4/tutorial4.4.compoundlist to the directory where the COSMO-RS database is downloaded, and select this file with Compounds → Add Compound(s). The tutorial4.4.compoundlist is a file with a list of compounds that is limited to the coumpounds needed in this example.
Select Compounds → Compounds Click on the left side 'Vanillin' Enter '355' in the 'Melting point:' field Enter '5.35' in the 'Δ_fusion H:' field Select Properties → Solubility Select 'Water' for the first component in 'Solvent' Select 'Vanillin' for the first 'Solute' Enter '298.15' without quotes in the 'Temperature from:' field in 'Kelvin' Enter '298.15' in the 'to:' field in 'Kelvin' Press 'Run'
Note that for the calculation of the solubility in units of mol/L and g/L the program uses information of the mass and COSMO volume of the solvent and solute. Especially the COSMO volume may not always be accurate.
In a similar way as we did the solubility calculation of Vanillin in Water, one can now calculate the solubility of Vanillin in several other solvents.
Select Properties → Solubility Select '1,2-Dichloroethane' for the first component in 'Solvent' Press 'Run' Do the same for the solvent '1-Propanol', 'Oxolane' (THF), 'Acetonitrile', '2-Butanone', 'Ethanol', 'Methanol', and 'Toluene'
These results can be put in a table.
| Solvent | experimental Solubility Vanillin (mol/L) [441] | calculated Solubility Vanillin (mol/L solution) | |
| 1 | 1,2-Dichloroethane | 1.175 | 2.65 |
| 2 | 1-Propanol | 1.820 | 3.12 |
| 3 | Oxolane | 3.594 | 4.73 |
| 4 | Acetonitrile | 2.360 | 3.30 |
| 5 | 2-Butanone | 2.138 | 4.10 |
| 6 | Ethanol | 2.470 | 3.56 |
| 7 | Methanol | 4.160 | 4.04 |
| 8 | Toluene | 0.302 | 0.99 |
| 9 | Water | 0.070 | 0.034 |
Compared to experiment there can easily be a factor of 2 deviation.
References
[441] Experimentally determined solubilities of Vanillin in some organic solvents were taken at July 15, 2010, from http://old.oru.edu/cccda/sl/solubility/allsolvents.php?solute=vanillin.
In this example a vapor-liquid diagram of Methanol and Hexane is calculated and compared to experiment.
First a binary mixture of Methanol and Hexane is calculated at a constant temperature of 333.15 Kelvin. Next this binary mixture is calculated at a constant pressure of of 1.01325 bar. Experimental pure compound properties are used.
Select Compounds → Compounds Click on the left side 'Methanol' Enter '0.845' in the 'Pure compound vapor pressure:' field Enter '333.15' in the 'at temperature:' field Click on the left side 'Hexane' Enter '0.77' in the 'Pure compound vapor pressure:' field Enter '333.15' in the 'at temperature:' field Select Properties → Binary Mixture VLE/LLE Select 'Methanol' for the first compound Select 'Hexane' for the second compound Enter '100' in the 'Number of mixtures' field Select Isotherm, isobar, flash point → isotherm Enter '333.15' for the 'Temperature' in 'Kelvin' Press 'Run' Select View → Graph X Axes → x1, y1 Select View → Graph Y Axes → total vapor pressure
In case of a miscibility gap there are two molar fractions x1 and x1', for which both compounds have the same activities. In the calculation one can plot the activity a1 versus a2. If there is a closed loop, there is a miscibility gap:
Select View → Graph X Axes → a1: x1*gamma Select View → Graph Y Axes → activities
Thus there is a calculated miscibility gap. The properties of the two immiscible liquid phases are given in the table: the calculated miscibility gap is between approximately x1 = 0.275 and x1' = 0.811, with a calculated total vapor pressure of approximately 1.48 bar. Within the miscibility gap, the liquid mixture consists of 2 immiscible liquid phases, one is Methanol-rich, the other Hexane-rich. Note, however, that within the miscibility gap the COSMO-RS calculation further incorrectly uses a forced 1 liquid-phase instead of 2 immiscible liquid phases. Also note that a pressure-maximum azeotrope is in the miscibility gap.
If the check box 'force 1 liquid phase within possible miscibility gap' is selected, then results will be shown also within the miscibility gap, with the unphysical conditions that the two liquids are forced to mix. If the check box 'force 1 liquid phase within possible miscibility gap' is deselected, then results will be shown in the graph and table only for those compositions of the mixture, which are outside of the miscibility gap.
Deselect the check box 'force 1 liquid phase within possible miscibility gap' Select View → Graph X Axes → x1: molar fraction 1 Select View → Graph Y Axes → excess energies
Select Compounds → Compounds Click on the left side 'Methanol' Enter '1.01325' in the 'Pure compound vapor pressure:' field Enter '337.8' in the 'at temperature:' field Click on the left side 'Hexane' Enter '1.01325' in the 'Pure compound vapor pressure:' field Enter '342' in the 'at temperature:' field Select Properties → Binary Mixture VLE/LLE Select 'Methanol' for the first compound Select 'Hexane' for the second compound Enter '100' in the 'Number of mixtures' field Select Isotherm, isobar, flash point → isobar Enter '1.01325' for the 'Pressure' in 'bar' Press 'Run' Select View → Graph X Axes → x1, y1 Select View → Graph Y Axes → temperature Select the check box 'force 1 liquid phase within possible miscibility gap'
There is a calculated miscibility gap between approximately x1 = 0.221 and x1' = 0.835, with a calculated temperature of approximately 323.3 Kelvin (50.1 °C). Within the miscibility gap, the liquid mixture consists of 2 immiscible liquid phases, one is Methanol-rich, the other Hexane-rich. Note that if the check box 'force 1 liquid phase within possible miscibility gap' is selected, within the miscibility gap the COSMO-RS calculation further incorrectly uses a forced 1 liquid-phase instead of 2 immiscible liquid phases. Also note that a temperature-minimum azeotrope is in the miscibility gap.
Experimental results for the Methanol-Hexane mixture were taken from Ref. [451]. These are compared with the calculated ones in the next graph. More experimental VLE data might also be found at Ref. [452].
References
[451] Wikipedia Hexane data page: http://en.wikipedia.org/wiki/Hexane_(data_page)
[452] CHERIC. Korea Thermophysical Properties Data Bank: Binary Vapor-Liquid Equilibrium Data
In this example infinite dilution activity coefficients of 31 organic compounds in Water are calculated and compared to experiment.
Remark: To only add the compounds needed in this example, one can also open a new COSMO-RS GUI window (SCM → COSMO-RS), copy the $ADFHOME/examples/crs/Tutorial4/tutorial4.6.compoundlist to the directory where the COSMO-RS database is downloaded, and select this file with Compounds → Add Compound(s). The tutorial4.6.compoundlist is a file with a list of compounds that is limited to the coumpounds needed in this example.
Select Properties → Activity coefficients Select 'Water' for the first component in Solvent Click the 'None' button next to 'Compounds' Click the search button below 'Solutes (infinite dilute)' to add some compounds: '1,1,1-Trichloroethane', '1,2-Dichloroethane', '1-Bromobutane', '1-Bromopropane', '1-Chlorobutane', '1-Chlorohexane', '1-Chloropentane', '1-Chloropropane', '1-Propoxypropane', '2-Bromobutane', '2-Bromopropane', '2-Chloropropane', 'Benzene', 'Bromoethane', 'Chloroform', 'Cumene', 'Dichloromethane', 'Diisopropyl_ether', 'Di-n-butyl_ether', 'Ethylbenzene', 'Ethyl_iodide', 'Mesitylene', 'm-Xylene', 'n-Butylbenzene', 'n-Propylbenzene', 'n-Propyl_iodide', 'o-Xylene', 'p-Xylene', 'Tetrachloroethene', 'Toluene', and 'Trichloroethene', Press 'Run'
These results can be put in a graph and a table. Experimental results taken from Ref. [461].
| Solute | experimental γ2∞ [461] | calculated γ2∞ | |
| 1 | Dichloromethane | 253 | 278 |
| 2 | Diisopropyl_ether | 628 | 2343 |
| 3 | 1,2-Dichloroethane | 641 | 976 |
| 4 | Bromoethane | 679 | 601 |
| 5 | Chloroform | 903 | 1474 |
| 6 | 2-Chloropropane | 1.48 103 | 1.42 103 |
| 7 | 1-Chloropropane | 1.75 103 | 1.73 103 |
| 8 | 2-Bromopropane | 2.09 103 | 1.77 103 |
| 9 | Ethyl_iodide | 2.19 103 | 1.19 103 |
| 10 | 1-Propoxypropane | 2.31 103 | 4.89 103 |
| 11 | Benzene | 2.48 103 | 2.07 103 |
| 12 | 1-Bromopropane | 2.86 103 | 2.36 103 |
| 13 | 1,1,1-Trichloroethane | 5.90 103 | 5.77 103 |
| 14 | 1-Chlorobutane | 7.61 103 | 7.25 103 |
| 15 | 2-Bromobutane | 8.32 103 | 7.81 103 |
| 16 | n-Propyl_iodide | 8.55 103 | 5.16 103 |
| 17 | Trichloroethene | 8.75 103 | 6.45 103 |
| 18 | Toluene | 9.19 103 | 8.14 103 |
| 19 | 1-Bromobutane | 1.22 104 | 0.93 104 |
| 20 | o-Xylene | 3.05 104 | 2.37 104 |
| 21 | 1-Chloropentane | 3.21 104 | 3.13 104 |
| 22 | Ethylbenzene | 3.27 104 | 3.00 104 |
| 23 | m-Xylene | 3.32 104 | 3.56 104 |
| 24 | p-Xylene | 3.33 104 | 3.30 104 |
| 25 | Tetrachloroethene | 3.60 104 | 3.81 104 |
| 26 | Di-n-butyl_ether | 4.72 104 | 8.95 104 |
| 27 | Cumene | 1.02 105 | 0.83 105 |
| 28 | Mesitylene | 1.17 105 | 1.49 105 |
| 29 | n-Propylbenzene | 1.36 105 | 1.24 105 |
| 30 | 1-Chlorohexane | 1.41 105 | 1.46 105 |
| 31 | n-Butylbenzene | 5.66 105 | 5.16 105 |
References
[461] J. Li, A.J. Dallas, D.I. Eikens, P.W. Carr, D.L. Bergmann, M.J. Hait, C.A. Eckert, Measurement of large infinite dilution activity coefficients of nonelectrolytes in water by inert gas stripping and gas chromatography, Analytical Chemistry 65 (1993), 3212
In Ref. [471] the ADF COSMO-RS parameters were fitted to 642 experimental values of properties for the set of 217 molecules from the supplementary material of Klamt et al. [472]. These properties are the ΔG of hydration, the vapor pressure, and the partition coefficients for Octanol/Water, Hexane/Water, Benzene/Water, and Ethoxyethane/Water. In this example all these properties will be calculated for the set of 217 compounds, and the results will be put in this table, and compared to the experimental data collected by Klamt. et al. in Ref. [471].
Open a new COSMO-RS GUI window (SCM → COSMO-RS). Copy the $ADFHOME/examples/crs/Tutorial4/tutorial4.7.compoundlist to the directory where the COSMO-RS database is downloaded. The tutorial4.7.compoundlist is a file with a list of compounds that is limited to the coumpounds needed in this example.
ΔG of hydration
Select Properties → Activity coefficients Select 'Water' for the first component in Solvent Enter '298.15' for 'Temperature Kelvin' Click the check box 'Use input density solvent (kg/L)' Enter '0.997' in the 'Use input density solvent (kg/L)' field Click the 'All' button next to 'Compounds' Press 'Run'
The results for the free energy of hydration (Δ G (kcal/mol) gas → solute) can be compared with experimental values, see this table and next graph.
Vapor pressure
Select Properties → Solvent Vapor Pressure Select 'Methane' for the first component in the Solvent Enter '298.15' for 'Temperature Kelvin' Press 'Run'
One could do this for each of the 217 compounds, and take the log10 of the calculated vapor pressure in bar. The results for the calculated vapor pressures can be compared with experimental values, see this table and next graph. Suggestion is to do this only for a few molecules.
Octanol-Water partition coefficients
Select Properties → Log Partition Coefficients Select '1-Octanol' for the first component in the solvent Select 'Water' for the second component in the solvent Enter '0.725' for the Mole fraction of 1-Octanol in phase 1 Enter '0.275' for the Mole fraction of Water in phase 1 Enter '1.0' for the Mole fraction of Water in phase 2 Click the check box 'Use input volume solvent phase 1/phase 2' Enter '6.766' in the 'Use input volume solvent phase 1/phase 2' field Enter '298.15' for the 'Temperature' in 'Kelvin' Click the 'All' button next to 'Solutes' Press 'Run'
The results for the Octanol-Water partition coefficients can be compared with experimental values, see this table and next graph.
Hexane-Water, Benzene-Water, and Ethoxyethane-Water partition coefficients
Select Properties → Solvent Log Partition Coefficients Select 'Hexane' for the first component in the solvent Select 'Water' for the second component in the solvent Enter '1.0' for the Mole fraction of Hexane in phase 1 Enter '0.0' for the Mole fraction of Water in phase 1 Enter '1.0' for the Mole fraction of Water in phase 2 Click the check box 'Use input volume solvent phase 1/phase 2' Enter '7.28' in the 'Use input volume solvent phase 1/phase 2' field Enter '298.15' for the 'Temperature' in 'Kelvin' Click the 'All' button next to 'Solutes' Press 'Run'
Select Properties → Solvent Log Partition Coefficients Select 'Benzene' for the first component in the solvent Select 'Water' for the second component in the solvent Enter '1.0' for the Mole fraction of Benzene in phase 1 Enter '1.0' for the Mole fraction of Water in phase 2 Click the check box 'Use input volume solvent phase 1/phase 2' Enter '4.93' in the 'Use input volume solvent phase 1/phase 2' field Enter '298.15' for the 'Temperature' in 'Kelvin' Click the 'All' button next to 'Solutes' Press 'Run'
Select Properties → Solvent Log Partition Coefficients Select 'Ethoxyethane' for the first component in the solvent Select 'Water' for the second component in the solvent Enter '1.0' for the Mole fraction of Ethoxyethane in phase 1 Enter '1.0' for the Mole fraction of Water in phase 2 Click the check box 'Use input volume solvent phase 1/phase 2' Enter '5.75' in the 'Use input volume solvent phase 1/phase 2' field Enter '298.15' for the 'Temperature' in 'Kelvin' Click the 'All' button next to 'Solutes' Press 'Run'
The results for the Hexane-Water, Benzene-Water, and Ethoxyethane-Water partition coefficients can be compared with experimental values, see this table and next graph.
References
[471] C.C. Pye, T. Ziegler, E. van Lenthe, J.N. Louwen, An implementation of the conductor-like screening model of solvation within the Amsterdam density functional package. Part II. COSMO for real solvents. Canadian Journal of Chemistry 87, 790 (2009)
[472] A. Klamt, V. Jonas, T. Bürger and J.C. Lohrenz, Refinement and Parametrization of COSMO-RS. Journal of Physical Chemistry A 102, 5074 (1998)
In the literature one can find several strategies to calculate pKa values. Some of these strategies involve the inclusion of explicit solvent molecules, in combination with a continuuum model, like COSMO. The strategies described in the next examples do not take into account explicit solvent molecules, only the continuum model is included.
In Refs. [511,512] one can find several strategies to calculate pKa values. The strategy described in this example does not take into account explicit solvent molecules, only the continuum model is included. Like in Ref. [512] an empirical fit is used to correlate calculated values with experimental pKa values, to account for systematic errors. The largest errors are probably present in the calculation of the solvation energies of the charged species. In these cases one probably should include explicit solvent molecules, however, that may introduce other sources of errors, and this will not be investigated here. The calculation of pKa values of acids (HA) and bases (B) is based on the following reaction model:
(acids) HA (aq, 1M) + H2O (l, 1M) → H3O+ (aq, 1M) + A- (aq, 1M) (5.1.1)
(bases) HB+ (aq, 1M) + H2O (l, 1M) → H3O+ (aq, 1M) + B (aq, 1M) (5.1.2)
The free energy of dissociation for acids and bases is calculated as
(acids) ΔG*diss = G(A-) - G(HA) + G(H3O+) - G(H2O) (5.1.3)
(bases) ΔG*diss = G(B) - G(HB+) + G(H3O+) - G(H2O) (5.1.4)
The * denotes a standard state of 1M (1 mol/L). The pKa can then be calculated as (see for example Refs. [511,512])
pKa = ΔG*diss/(RT ln(10)) - 1.74 (5.1.5)
At T = 298.15, 1/(RT ln(10)) = 0.733 mol/kcal. The term -1.74 is to correct for the standard state of liquid water, which is 55 mol/L.
Emperical fit
Like in Ref. [512] instead of this equation (5.1.5), a linear fit has been made by correlating the calculated ΔG*diss values with experimental pKa values, to account for systematic errors that are present in this method. For acids and bases a different empirical adjusted equation will be used, optimized for the ADF COSMO-RS implementation:
(acids) pKa = 0.62 ΔG*diss/(RT ln(10)) + 2.10 (5.1.6)
(bases) pKa = 0.67 ΔG*diss/(RT ln(10)) - 2.00 (5.1.7)
These fitted parameters are not so far from the fitted parameters in Ref. [512]. Zero-point vibrational energies have not been taken into account in the calculation of the free energy of dissociation. H3O+ (Hydronium ion) is the conjugate acid of Water. Molecules can have two or more equivalent sites for protonation or deprotonation are also not taken into account, which can have an effect on the pKa value. However, such effects are not taken into account here. Like for neutral compounds, one should optimize the anions and cations in the gas phase, and use this geometry also in the COSMO calculation. It is important to choose the lowest energy conformer. In the example below the molecules have a single relevant conformation for the protonated and deprotonated form.
Acids
Copy the .coskf files which are listed in $ADFHOME/examples/crs/Tutorial5/tutorial5.1_acid.compoundlist and the file tutorial5.1_acid.compoundlist the directory Tutorial. The tutorial5.1_acid.compoundlist is a file with a list of acids and their conjugate bases that is limited to the coumpounds needed in this example. In these .coskf files already the correct number of ring atoms is included.
Open a new COSMO-RS GUI window Add the compounds listed in tutorial5.1_acid.compoundlist Select Properties → Activity coefficients Select 'Water' for the first component in Solvent Enter '298.15' for 'Temperature Kelvin' Click the 'All' button next to 'Compounds' Press 'Run'
Using equation 5.1.6 (pKa = 0.62*0.733*ΔGdiss+2.10), with ΔGdiss = G(conjugate_base_acid) - G(acid) + G(conjugate_acid_Water) - G(Water) in kcal/mol, results can be put in a table.
| Acid | experimental pKa [512] | calculated pKa | |
| 1 | Methanol | 15.5 | 16.27 |
| 2 | Ethanol | 15.9 | 15.97 |
| 3 | Phenol | 9.82 | 10.32 |
| 4 | Acetic acid | 4.75 | 3.64 |
| 5 | Benzoic acid | 4.27 | 4.28 |
Bases
Copy the .coskf files which are listed in $ADFHOME/examples/crs/Tutorial5/tutorial5.1_base.compoundlist and the file tutorial5.1_base.compoundlist the directory Tutorial. The tutorial5.1_base.compoundlist is a file with a list of bases and their conjugate acids that is limited to the coumpounds needed in this example. In these .coskf files already the correct number of ring atoms is included.
Open a new COSMO-RS GUI window Add the compounds listed in tutorial5.1_base.compoundlist Select Properties → Activity coefficients Select 'Water' for the first component in Solvent Enter '298.15' for 'Temperature Kelvin' Click the 'All' button next to 'Compounds' Press 'Run'
Using equation 5.1.7 ((pKa = 0.67*0.733*ΔGdiss-2.00), with ΔGdiss = G(base) - G(conjugate_acid_base) + G(conjugate_acid_Water) - G(Water) in kcal/mol, results can be put in a table.
| Base | experimental pKa [512] | calculated pKa | |
| 1 | Aniline | 4.6 | 5.75 |
| 2 | 1H-Imidazole | 7 | 5.65 |
| 3 | Pyrazine | 0.7 | 0.85 |
| 4 | Pyrazole | 2.5 | 2.55 |
| 5 | Pyridine | 5.14 | 4.62 |
| 6 | Quinoline | 4.80 | 4.43 |
| 7 | Guanidine | 13.8 | 14.09 |
References
[511] J. Ho and M.L. Coote, A universal approach for continuum solvent pKa calculations: are we there yet?, Theoretical Chemistry Accounts 125, 3 (2010)
[512] F. Eckert, M. Diedenhofen, and A. Klamt, Towards a first principles prediction of pKa: COSMO-RS and the cluster-continuum approach, Molecular Physics 108, 229 (2010)
The method described in this example is based on one of the strategies in Ref. [521]. This method uses the experimental pKa value of a reference compound, experimental gas phase deprotonation energies, and COSMO-RS solvation free energies. A suitable reference compound (HRef) should be chosen, which is similar to the actual compound (HA) one is interested in. For example, the deprotonation reaction could be similar in both compounds. The calculation of pKa values is based on the following reaction model:
| HA(aq, 1M) | + | Ref -(aq, 1M) | ΔG*soln → |
HRef(aq, 1M) | + | A-(aq, 1M) |
| ↓ -ΔG*solv (HA) | ↓ -ΔG*solv (Ref -) | ↑ ΔG*solv (HRef) | ↑ ΔG*solv (A -) | |||
| HA(g, 1M) | + | Ref -(g, 1M) | → ΔG*gas |
HRef(g, 1M) | + | A-(g, 1M) |
The * denotes a standard state of 1M (1 mol/L). The free energy of solvation is then calculated as
ΔG*soln = ΔG*gas + ΔG*solv(HRef) + ΔG*solv(A-) - ΔG*solv(HA) - ΔG*solv(Ref -)
The pKa can then be calculated as (see for example Ref. [521])
pKa = ΔG*soln/(RT ln(10)) + pKa(HRef)
The success of this method relies on the availability of a suitable reference compound with an accurately known experimental pKa value. Instead of experimental gas phase deprotonation energies one might calculate the gas phase reaction free energy ΔG*gas, using DFT or some high level ab initio method.
Example Ethanol
In this example the pKa value of Ethanol will be calculated. As reference compound Methanol is chosen, which has an experimental pKa value of 15.5. Experimental deprotonation energies are taken from Ref. [521, supporting information], for Methanol this is ΔrG0 = 1569 kJ/mol, and for Ethanol ΔrG0 = 1555 kJ/mol, see also, for example, Ref. [522]. The COSMO-RS solvation free energies of Methanol, Methoxide (the conjugate base of Methanol), Ethanol, and Ethoxide (the conjugate base of Ethanol) have already been calculated in the previous tutorial 5.1. At T = 298.15, 1/(RT ln(10)) = 0.733 mol/kcal. The free energy of solvation (kcal/mol) and the pKa of Ethanol are then calculated as
ΔG*soln = (1555-1569)/4.184 - 4.97 - 86.56 + 4.79 + 90.14 = 0.05 kcal/mol
pKa (Ethanol) = 0.733*0.05 + 15.5 ≈ 15.5
The calculated pKa of Ethanol of 15.5 is close to the experimental value of 15.9.
References
[521] J. Ho and M.L. Coote, A universal approach for continuum solvent pKa calculations: are we there yet?, Theoretical Chemistry Accounts 125, 3 (2010)
[522] NIST Chemistry WebBook
