Developing new reactions

This definitely a newbe question, I have finished going through years of theoretical OCHEM, but have subpar lab experimentation (other than academically assigned and also a few dabs here and there).

I wish to become proficient in devising my own experiments. We as a science community understand chemistry a lot better than they did 50, 60, 70 years ago when many of the first chemical synthesis books were first published. Let's not forget that books in the last decades have a hybrid of old methods intercalated with a few newer ones.

Without denial, the key to becoming a good chemist is practice; the practicality aspect of my experience is not overlooked, but the keys to success is to have the right tools available, which brings me to my point.

When devising an experiment that may or may not be in an Organic Synthesis "Bible", for example, the following 2 reactions



Here were my steps in thinking:
1) I knew I wanted markovnikov addition for alcohols and bromine. So Br2 in H2O should give in theory the chosen product. However Br2 being so unhindered and reactive would give more side products (ref 1) so NBS is better suitable, which reaction A is based on (ref 2).
2) the equivalents of NBS and alkadiene should(?) be 2:1, basic stoichometry
3) DMSO/H2O, because I have DMSO available, though I am thinking about THF since it has a lower bp (thinking about product purification).
4) the solvent chosen is water (which in this reaction acts as a nucleophile and as a solvent for the product, and the alkene solubility is arranged by the THF/DMSO/DMS etc.) - even when reading about solvents and reagents (ref 3), I do not fully comprehend which optimal solvent to be chosen.
5) purification of diol, unless the product is a solid, seems I doubt, may be a little complicated: getting rid of NBS and DMSO/THF (which are good solvents). Hypothetically if the next step is protection of alcohols, could I just add pyridine and Ac2O, or this is considered bad practice and doubtful results?


side question: how is the input heat of the reaction determined? trial and error or a computational way of doing it?
Thanks for reading, it was a lot, but there's so many questions that as a newbe, I've come across that no book really answered, and often professors just give textbook type of answers.


References

1. Wiki: Halohydrin formation rxn
2.A guide to solvents and reagents
3. NBS


[Edited on 5-10-2010 by andre178]

There are reported HOBr additions in DMSO/H2O. However, I also don't see this reaction working effectively. NBS forms hypobromite, which reacts with alkenes in a Markovnikov fashion, so you'll get a bis-bromohydrin with bromine at the terminal positions. You'd need to use an excess of NBS.

I don't want to be a downer, but chemistry is a very well developed field at this point. Developing new reactions is truly an immense challenge. The chemistry 'bible' is the primary literature, and this is ridiculously large collection of work. Don't fool yourself, chemists even 100 years ago were quite knowledgeable - you would be surprised. The likelihood of you being able to develop a 'new' reaction at home is slim to none. Professional researchers devote an unbelievable amount of thought, time and effort to developing even a single new methodology or reaction. You're better off buying more advanced textbooks (March's Organic Chemistry, Zweifel & Nantz - Modern Organic Synthesis) and building on your knowledge from there. As for what you'll be able to achieve at home, this is dependent on how much you're willing to break away from the 'otc' mantra. You really need to convince yourself that it's okay to purchase intermediates and reactants from chemical suppliers if you want to do some clever chemistry. I've got plenty of legal ideas, so surely you do too

Andre178, what you are asking is about designing a new synthesis, not developing a new reaction. The reaction you ask about is way over a century old. It is called electrophilic addition on alkenes, more specifically it is about bromohydroxylation.

What you need to do is to determine the most likely products first. For this you use mechanistics and primary literature. Thus, you need to check the mechanism of bromohydroxylation (electrophilic addition on alkenes) and search the literature for this specific or related synthesis/method/product. For example, it helps if you first make a chart of all possible bromohydroxylation products of your substrate and then rule out the unlikely ones (those not conforming the regioselectivity, stereoselectivity, statistics, ring size, Baldwin's rules and so on). Bellow is an example of few such products, but they do not include the non-bromohydroxylation products, the ones breaking the (anti) stereoselectivity of the nucleophilic attack on the bromonium ion and few other things (it would be too much to include everything, just to rule out immediately). They do however include products that are unlikely due to incorrect regioselectivity (c, d, e, f, g) and/or products that are unlikely due ring size thermodinamics (j, k, l, m). Regioselectivety of reactions is often non-specific, that being, some reactions commonly give mixtures of regioisomers, so products derived from lack of regioselectivity should not be immediately ruled out without consideration.
You see that just applying a few basic rules you rule out a whole bunch of products. Applying statistics gives you the information that the amount of a formed will be more or less equal to b, as the diastereoselectivity is expected to be low, due to the lack of steric influences. Same goes for h and i diastereomers. Applying the knowledge of thermodinamics and Baldwin rules will tell you that an intramolecular reaction is more likely than a intermolecular one, thus giving you the idea that the products h and i should form in higher amounts than a and b, but this is obviously highly dependent on the concentration of H2O and method used. Furthermore, the awareness that you are dealing with a symmetric reaction will give you the information that all the products formed will be racemic.
This is only about the reaction and you also need to consider the method. If the use of NBS in aqueous DMSO is not yet described for simple terminal alkenes like the one in question, you need to think about the possible side reactions. Haloimides react with DMSO (with the rate highly depending on the haloimide used, the pH, temperature and other parameters), so if the desired reaction is slow, some NBS will be consumed by this side reaction. Carbocations get oxidized by DMSO to give ketones/aldehydes. Since bromonium ions are in equilibrium with the appropriate carbocation, you can only hope the nucleophilic addition of H2O on the bromonium ion is faster than the nucleophilic addition of DMSO, else this becomes another possible source of side products. Side reactions consuming/reducing NBS give bromide as one of the side products and this can interfere with the bromohydroxylation. It can react with NBS to give Br2 which can add to the alkene or it can nucleophilicaly add on the bromonium intermediate, in both cases leading to 1,2-dibromoalkanes instead of bromohydrines. So the choice of bromohydroxylation method is important if you want to reduce the amount of 1,2-dibromoalkanes side products.

This were just examples of a few things to consider and there are more!
If you ever graduated from organic chemistry you should have been thought all this already (if not, make a reclamation of your schooling!). What you need when planing a synthesis is being able of analytical and divergent thinking, and being able to perform primary literature searches. The first is highly dependent of the culture you were raised in, the schooling system and your ideological background. Unfortunately, in the last few decades most educational systems in the world are supporting convergent thinking and suppressing divergent thinking, so schooling can often be detrimental for those pupils who would want to become researchers. The ability of performing primary literature searches depends on you learnt skills as well as access to abstracts databases and literature, but unfortunately it is also highly dependent on your personality.

Quote: Originally posted by Arrhenius

Ouch.. I'm going to disagree on any formation of 7 and 8 membered ring ethers by Sn2. May be 'allowed' by Baldwin rules, but highly unlikely to form.