# How are solvents chosen in organic reactions?

Just reading through organic chemistry book (Modern Projects in Organic Chemistry by Mohrig, et al.), and it seems like the choice of solvent is usually arbitrary.

One of the experiments designed uses hypochlorite ion as the oxidizer and acetone as a solvent to oxidize 4-tert-butylcyclohexanol, but it isn't clear why this solvent is chosen.

I'm curious if there is a list that gives the suggested solvents for different kinds of reactions, as well as some insight for this particular experiment.

There is no list per se, but there is certainly a process to determine appropriate solvents for a given reaction. For the specific reaction you mention, here are some criteria you would consider to select a solvent:

1. Because this is an oxidation reaction, choose a solvent that cannot be oxidized, known as an aprotic solvent.
2. The solvent needs to dissolve the reagents in the reaction: hypochlorite and 4-tert-butylcyclohexanol. Hypochlorite is a polar molecule, and 4-tert-butylcyclohexanol is a little polar thanks to the alcohol group. The alcohol group will be oxidized to a ketone by the reaction, so the solvent should be able to dissolve that reasonably well. Conclusion: choose a moderately polar solvent.

So a polar aprotic solvent is needed for this reaction. There are two more considerations: cost and safety

Obviously, the cheapest solvent with the lowest potential for harm should be selected.

A list of some common polar aprotic solvents for consideration might include:

1. tetrahydrofuran
2. N,N-dimethylformamide
3. acetonitrile
4. dimethyl sulfoxide
5. acetone

Tetrahydrofuran could be a reasonable choice, but it forms peroxides over time, which are quite explosive.

N,N-dimethylformamide has some health concerns: GHS07 and GHS08

Acetonitrile also falls into the GHS07 category: it causes cyanide to form in the body at high concentrations.

Dimethyl sulfoxide does not have health or safety concerns.

Acetone also falls into the GHS07 category: it can cause nausea and irritation at high exposure levels.

The health concerns of acetonitrile, dimethyl sulfoxide, and acetone are largely not going to pose a hazard for a laboratory experiment. All three of these can likely be found in any organic teaching lab.

Cost comparison, from Sigma Aldrich:

Acetonitrile: 2 L = $172.50 Dimethyl sulfoxide: 2 L =$208.50

Acetone: 2 L = $94.80 Therefore, choose acetone for this reaction! • Another reason: DMSO is incredibly difficult to get rid of. – orthocresol Oct 6 '16 at 7:38 • To be precise, "cannot be oxidized" and "aprotic" are two rather distinct things. (Though I will admit for many common solvents they do line up.) e.g. acetaldehyde (if for some bizzare reason you tried to use it as a solvent) would be aprotic, but is very easy to oxidize. – R.M. Oct 6 '16 at 14:09 • Also DMSO may be safe, but it can readily transport reagents/reaction products through skin (hence its use in percutaneous drug delivery); the solute may be less safe. – Chris H Oct 6 '16 at 16:00 •$\text{aprotic} \ne \text{non-oxidiseable}\$. Most importantly, acetone is protic! – Jan Oct 6 '16 at 20:15
• I wasn't expecting the answer to get into the economics of the chemical catalog, I wonder if this isn't the biggest factor of all? – Ryan Ward Oct 7 '16 at 19:58

I did organic chemistry in the lab for 9 years, and never once was there such a list. There are general rules for protic vs. aprotic and polar vs. non-polar, but a lot of times, when a reaction is published, the authors did a study of different kind of solvents and selected the "best" based on yield, reaction rate, ease of purification, or selectivity.

If you consider YANR (your average name reaction), the scientists who published said YANR will typically have done a set of screenings, including a ‘solvent screen’. That means, they will have run said reaction in a number of different common laboratory solvents and compared the yields. The table will then look something like this (disclaimer: all values in the following table made up on the spot):

$$\begin{array}{cc}\hline \text{Solvent} & \text{yield} \\ \hline \text{hexane} & 55~\%\\ \text{DMF} & 72~\%\\ \ce{CH2Cl2} & 67~\%\\ \ce{EtOAc} & 59~\%\\ \text{toluene} & 85~\%\\ \textbf{acetone} & 89~\%\\ \hline\end{array}$$

And then they will proceed to note that acetone is their solvent of choice for YANR and proceed to further screenings (additives, catalysts, bases …) before finally recording a full substrate scope table. You may have noticed that this is trial and error.

For those reactions that have been around for a hundred years, nobody really did that optimisation. Instead, the solvent is used that you always have used. Occasionally, a paper turns up in a medium-tier journal that the yield of the reaction can be improved by modifying the conditions in the following ways, including the new solvent FancyDissolve™. Rarely, a top-tier paper turns up doubling the yield and improving the selectivity by orders of magnitude because they chose dichloromethane rather than hexane which never occurred to anybody.

That is the secret behind selecting solvents.

There are many factors. You have to consider the solubility of reactants since the reactants have to mix to react better. There is a solubility chart for several organic solvents. A lab that I used to work in had one pasted on the wall.

Another factor is whether the solvent interferes with the reaction. In other words, you want to minimize side products. An example of this are Grignard reagents: you do not want protic solvents when working with them. Similarly, you do not want your solvent to chelate/react with a catalyst (unless it is a reaction that requires a chelate).

Another one is availabity. This can be cost availabilty or information availability. If there are studies that are similar to the experiment that you are conducting, you will probably use the solvents similar to he ones that were used in previous studies. Cost may play a factor as well. Why pay extra for an exotic solvent when a cheap common solvent yields ok results? (which sounds like a crime in science...but funding is not always easy to get).

• Whether or not being satisfied with OK results depends on what your goal is. If making the stuff is your goal, OK probably isn't good enough. But if making the stuff is only a means to an end then OK is probably fine. – David Richerby Oct 6 '16 at 9:01