Traditional medicines can be interesting things. Some don't work at all despite being widely used, but some, like the bark and leaves of willow (Salix), work very well and have been effectively used for millennia—there are written records from 500 BC referring to its use. More recently (the 1820s) the active ingredient, salicylic acid, was produced, then in the 1890s acetylsalicylic acid, what we know as modern aspirin or ASA, was created. Aspirin no longer comes from plants as willow trees can't grow fast enough to sate the world's appetite for painkillers, but is now synthesized from phenols.
Another traditional painkiller, a milkwood (Tabernaemontana) has been under investigation for several years now. According to the studies, it contains a mixture of several things, including compounds in the class of opioids (a painkiller type which tends to have undesirable side effects and which causes addiction) and conolidine, among many others.
The conolidine and other compounds were isolated and identified in 2004, but conolidine couldn't be properly studied at that time because they only managed to get a 0.00014% yield when purifying it out of the plant. In May 2011, a team of researchers from the Scripps Research Institute announced that they had not only managed to synthesize conolidine, they had also tested it on mice and found out that it had painkiller effects as strong as morphine, but without any of morphine's adverse side effects.
The synthesis they developed is a nine-step process with an overall 18% yield, much better than the 0.00014% yield achieved in purifying from the plant source.
The actual nine steps are all using organic chemistry, which I haven't worked with since that one class in second year university. It's all very exotic to me, because most of my work since graduation has been in aqueous inorganic solution sorts of things, not structures with multiple rings with stuff hanging off them. (Sorry, organic chemists. I'm too out of my depth here to say anything more intelligent than that about the structure.) I had initially hoped to talk about the chemistry of the synthesis, because it looks fascinating. Despite my propensity for diving into the deep end of a topic, organic chemistry's pool of knowledge is a whole lot too deep to study enough in a week or two after work to make a sensible article out of this paper. So I'm going to talk about the separations they used instead. That's more of an engineering topic anyway.
According to the supplementary information paper, they used only four basic separation principles, plus a couple that were not specified enough for me to classify them:
- Vacuum separation
- Liquid - liquid separation
Filtration is probably the easiest separation principle to explain. You take a mixture of liquids and solids, and pour it on a paper or cloth or other surface with small holes in it. The liquids go through and the solids don't. Industrially, it's sometimes a bit more interesting than that: sometimes the solids plug the holes in the filter media and then the liquids can't get through either, leaving you with a sloppy mess in the filter. Sometimes the solids are so tiny that they go right through with the liquid, messing with equipment or chemistry downstream that relies on a certain low level of solids. Or, if the solids are the product, then you lose it and your productivity suffers. There are all kinds of pretreatments, from coatings on the media to coagulants that clump together the solids, to make things filter faster and to a denser, drier filter cake.
Vacuum separation takes advantage of the fact that many compounds that otherwise mix very well have different vapour pressures. It's complementary to distillation; vapour pressure varies with temperature, and every chemical has a different one. In distillation, you heat the mixture, and the compound with the higher vapour pressure at the lower temperature leaves solution first to be carried away in the gas phase. This is used to produce everything from gasoline to whiskey. In vacuum separation, you lower the pressure over the solution, and the compound with the highest vapour pressure is carried away as a gas the fastest. Vacuum separation is especially useful if the compounds could be damaged by heat, or if vacuum is cheaper than heat. A lot of compounds have such vanishingly small vapour pressures that they stay put and form a residue if you keep the separation going until it stops changing: sometimes a solid, if it precipitates; sometimes a liquid.
Most of the vacuum separations listed in this process seem to be a complete solvent removal, leaving behind only the solutes. One named a compound removed by vacuum removal, and that would probably have been more of a selective separation: compounds remained behind that could be removed by vacuum but were not. This requires more careful control than complete solvent removal.
Liquid-liquid separations are really neat. The basic principle at work here is twofold: one, that sometimes two liquids don't mix with each other (oil and vinegar is the familiar example of immiscible liquids); and two, that things which are soluble in water are sometimes also soluble in those other liquids that don't mix with water.
As with vapour pressure mentioned above, every compound will have its own solubility in various solvents. Something might be twice as soluble in oil as it is in water, for example. In that case, if you mixed that solute in a jar with both oil and water and shook it well, after the oil and water separated there would be twice as much of the solute in the oil as there is in the water. If you pour off the oil from the mixture, then add some fresh oil without any solute present to the water and mix it, the one-third of the solute that had been in the water will split itself so that there is again twice as much in the oil as in the water, leaving 1/9th of the original amount of solute in the water. If you repeat that process a couple more times, you can get almost all of the solute out of the water by washing it with oil repeatedly.
The conolidine synthesis used diethyl ether, a good general purpose solvent for organic stuff, as the second phase to get their product out of water in most of the liquid-liquid separation steps. When the other solutes in the water solution are less soluble in diethyl ether, they tend to stay in the water while the product moves over to the ether side, letting you separate the product from the rest of the stuff in solution.
Chromatography is also an interesting one. This works on the principle that small molecules travel faster than large molecules. Or, molecules that are less attracted to the column material move faster than molecules that are more attracted, depending on the system. You may remember doing a simple paper chromatography experiment as a kid in science class, with ink, paper towel, and water. As the water is drawn up the paper towel, it separates black ink into a rainbow of colours, because some of the colours that go into making black ink travel more easily in the water/paper towel chromatograph. Once the paper towel had dried, you could cut apart the different colours to separate them.
In chromatography a carrier liquid or gas pushes the sample through until it comes out the other end. Because the different components travel at different speeds, each one will come out of the chromatograph at a different time, and it is possible to capture each one separately. The team who developed the conolidine synthesis found that they could get the stage product at better than 95% purity using this method, which is probably why it shows up five times in their nine-step process.
There's a lot of work to be done before this ever comes close to the pharmacy, but now that we know it does in fact act as a painkiller in mice, there will be further trials. It will be interesting to see what the commercial scale process looks like if this is eventually approved.