Mini Project

This is the part of the lab where you can be more independent. It is important that you check that the project is appropriate with Dr Hunt first.

The mini project is 50% of the final mark so you should be doing a substantial amount of work.

The purpose of this part of the project is to use the techniques you have learned on an interesting problem. You are carrying out "original" research and so justifying the particular calculations you carry out (ie what information you want to obtain), analysing your results and then forming a good well justified argument (even if it is negative) are more important than getting a definative answer.

important Check your progress regularly with a demonstrator or Dr. Hunt to ensure you are on the right track, and have not done too much or too little. In open ended projects it can be difficult to get the balance between number of calcuations vs the amount of interpretation correct (it should be about 50:50)

important Ask us questions! Early rather than late. You will get far more out of the project, and even have fun, if you have a clear idea of what you are doing. If you have problems, ask about why the problem has arisen, don't just try to fix it and move on.

Some helpful hints

Where to get ideas for the mini project:

Carrying out a project that investigates and allows you to compare electronic or structural effects are good.

• if there is something you find particularly interesting from your inorganic lectures. Taking material from third year lectures is highly challenging, as they tend to cover molecules that exhibit unusual bonding and are exciting! However, as beginers I highly suggest something a bit simpler for your first foray into computational chemistry and bonding analysis, something from first or second year is a better option.
• Do NOT choose a system with transition metals, unpaired electrons, or a radical ... these require special techniques which we haven't covered in the lab.
• I highly recommend looking through "Inorganic Chemistry" by Housecroft and Sharpe for inspiration. Most of the molecules examined in this text are within computational reach and the text often highlights interesting bonding, structural or electronic properties that can be used as the starting point for a project.
• you can base a project around a previous synthetic lab from either year 1 or 2
• equally if you have been reading the literature and find a particularly fascinating paper to topic you wish to explore this is encouraged
• don't choose a molecule with extensive or bulky ligands, often such ligands are used so that the synthetic chemist can crystalise their product ... we don't need them! And they make the calculation much, much longer ... you won't have time to finish.

Quality and quanitity for your project:

• Try to choose a group of molecules with between 4-10 atoms (not including H's), you should expect to carry out several geometry optimisations and to compare structures across aproximately 3-4 molecules.
• You should choose a method and basis set and justify its use. Smaller molecules should be calculated at a higher level than larger molecules, since you don't know that much about the different methods and basis sets yet, a good level to end up would be something like B3LYP and 6-31G(d) but you may want to try MP2 and 6-311G(d,p) for smaller molecules and more accurate calculations
• Do NOT use chem3bio to set up your structure or molecular mechanics to pre-optimise it, these methods do not work well with inorganic molecules, they are made to look at biological molecules. If you are unsure of how to build your molecule read the help menue in gaussview, if you still have problems ask a demonstrator.
• you should carry out an analysis of the geometry of your compounds, comparing key features. However, keep this section short as the main emphasis should be on the electronic structure.
• you should consider the vibrational spectrum of your molecules, analysing the spectrum of one moleule in detail, or comparing key parts of the spectra for your series of molecules
• you should examine the electronic structure analysing the MOs of one molecule in some detail, this means documenting ALL the occupied and non-core MOs, considering only the HOMO and LUMO is not sufficient. Do not examine unoccupied orbitals except for the 1 and 2LUMO, LUMO orbitals are not well represented by the methods you are using. You should pick 5-10 MOs to discuss in detail (considering their LCAO make-up, bonding characteristis and impact on bonding within the molecule), and if you have a small molecule produce a MO diagram.
• you should carry out an NBO analysis. Then make some comparisons and comments, perhaps comparing the changes that occur on substituting heaver elements, or electron donating or withdrawing ligands.
• you can compare energies, but only relative energies of molecules or reactions with exactly the same number and type of atoms are valid! This has been a problem area, so if you want to compare energies, please talk with a demonstrator first.
• use your data! at least 50% of the project should be centered around interpretation, compare and analyse your data, can you find a justification or interpreation for trends? Are you able to rationalize, using this information, some key property or chemistry of the compound?
• it is important that you can show that you have done the calculations, so make sure you lodge the final optimised and frequency calcultaions in the digital repository, and note the link on your wiki.

Setting out your project:

• You should have a clear idea of what you want to know before you start! For example "I want to know if x is more stable than y, and to be able to rationalise this ordering of stability" or "I want to investigate the electronic and energetic reasons for the instability (or stability) of x".
• From this you will be able to work out which types of calculations you need to carry out. Computational chemistry is a tool, you want to employ it to solve a problem or answer a question.
• You could try writing out three questions you want to answer in relation to your molecule or molecules of interest.

Important pointers which you ignore at your own peril!:

• The examples you have run in the first part of the lab have been tuned to run very quickly, you can expect the jobs for your project to take between 2-6hrs each to run.
• Experience has shown that students don't know when a job is running for too long hence it is important that in every optimisation the input file should be edited to include OPT(MAXCYCLE=50). Get a demonstrator to help you if you need it. If a job does not converge within 50 cycles you MUST get a demonstrator or myself look at it. If the job has not converged, you will most likely get an error when trying to open it in guassview.
• Your starting structure must be sensible, don't have atom-atom interactions that don't appear in the molecule. Think of the computational methods employed here as a refinement tool, they are not a replacement for you creating a structure carefully and with thought. For example, you should also find literature or sensible estimates for the bond distances of "heavy" atom contacts and set them explicitly using the bond tool.
• Always optimise first at a low level of calculation (as shown in the earlier section), a good point is B3LYP and 3-21G or LANL2MB.
• EVERY "completed" optimisation should be followed by a frequency analysis to confirm you have a minimum.