The Half Decent Pharmaceutical Chemistry Blog

Pharmaceutical chemistry is back! So far, I've described syntheses of rather banal molecules (although aspirin is obviously the cornerstone of any pharmaceutical chemist), but today, I'm going to discuss how we prepared a proper drug: tacrine.

What's more, there were rumours circulating during the lab course that this particular way of synthesizing this cholinesterase inhibitor was based on an article written by one of the two professors who looked after us. Whether this true true or not, this preparation is alternative to that I studied in one of my Pharma Chem courses.

How can I sum up this drug? It was the very first molecule to be used in the treatment of Alzheimer's disease. Tacrine is an indirect-acting cholinomimetic and (reversibly) binds to the active site of cholinesterase, through a cation-pi interaction and a hydrogen bond, the enzyme which performs the degradation of the neurotransmitter acetylcholine, hence, blocking it.

Cholinesterase was thought to play a major role in the development of Alzheimer's disease.
Although, nowadays, we have realised the progressive degeneration of cholinergic neurons isn't the sole phenomenon occurring in this pathology and tacrine isn't even the cholinesterase inhibitor of choice (due to adverse effects and pharmacokinetics) any more, this drug led the way in the development of many drugs and therapies to tackle this neurodegenerative disease. So, it does deserve a lot of respect.

0.5966 g of 2-aminobenzonitrile were put in a 100 mL round-bottom flask. Then, 1.733 g of zinc chloride powder, which catalyzes the reaction, were added.

The second reagent is ciclohexanone: 0.54 mL were introduced, drop by drop, in the flask.

Then, it was time to heat the mixture and let the reaction occur: the flask was fixed, closed, put in a silicon oil bath and we set the hot plate stirrer to 210°C for a couple of hours.

That was the end, because the synthesis covered two days (we were doing more than a synthesis a day). So, we went home and, on the next day, before we arrived, all the hot plates were turned on again.



At this point, a tough, crunchy, thick and brownish precipitate was stuck onto the inner surface of the flask. To quench the reaction and recover the product, the flask first allowed to cool to room temperature and, then, approximately 10 mL of water and 10 mL of a 5M NaOH solution were added: in these conditions, water quenches the reaction, inactivating zinc chloride.

Predictably, that wasn't sufficient to remove the precipitate from the bottom of the flask, so I had to spend some time removing it with a glass rod, scratching hard.

At last, all the product was dissolved and we stirred the white solution we ended up with for 30 minutes.



The resulting white powder was filtered with a Büchner funnel and suspended in ethanol (using four times the volume of the product, although, theoretically, twice should have been enough). Given that our aim is to synthesize tacrine as its hydrochloride, we added concentrated HCl until the pH of the (orange or red) suspension was undoubtedly acid.

The beaker were all these operations had been performed was carefully placed in a drawer and the yellowish crystal of tacrine hydrochloride were allowed to form and precipitate for approximately 20 minutes.

The product was filtered and we calculated the yield and determine its purity.



Weirdly, no one told us what the yield was supposed to be, so I can't say whether our 52.98% (0.628 g) was a huge success or a pathetic result. Moreover, a PhD student, who was helping the professor, told us that, in her opinion, the scheme we were following was wrong...

Nevertheless, the melting point (282-285°C) and the IR spectrum (I'm sorry for the nujol, Uncle Al) proved the product was at least pure.



A mixture (9:1:0.1) of dichloromethane/methanol/ammonia (to help bases to run, according to their properties) was the mobile phase for a TLC we did to get a better idea of the purity of our product (which was dissolved in hot methanol). The layer was, in the end, exposed to UV light and the stains were identified (picture coming soon).

Well, not exactly the opposite of yesterday's reaction, but the product is camphor and we start from the endo-isomer of isoborneol: borneol.

It would have been cool to first reduce camphor to isoborneol and then oxidize it back to check how much you lost in this pathway.
Unfortunately, while the reduction was performed on April 27, the oxidation took place two days ago, on Friday, once we had already thrown away our precious isoborneol.

Nevertheless, my aim in this post is to do something radically different from anything you've ever read on any other prestigious organic chemistry blog, as well as in this section of this one.

That's because this oxidation was done in a particular situation: as I said, the lab course is now over and this synthesis was selected by our professor as a sort of final test. For example, this was the first occasion when we worked on our own, instead of in pairs as usual.

I agree with you that this doesn't look like a particularly complicated reaction and, indeed, it isn't. Still, the major problem was the lack of time.

To work on our own, the students had to be split in two groups: a first half working in the morning and the other in the afternoon. In a nutshell, this meant we had to complete the reaction, determine the yield, check how pure your product actually is (with the help of TLC, melting point and IR spectra) and wash all the bloody glassware you had utilized for the person who would used it. All this in three hours and a half.

That gave the whole thing a race-against-the-clock effect, which I hope to properly recreate in this article.

So, if you have to go to the loo, do it now and, then, let the race begin!

9:00 am Before we could begin to crack borneol, our professor wanted to unveil the final test and discuss the uses of camphor, its properties...

9:45 am Eventually, the lecture was over, and we were finally in the lab. Approximately 4 g of sodium bichromate were dissolved in 16 mL of water (the sodium salt is more soluble in water than the potassium one). Then, 3.2 mL of concentrated sulphuric acid were added drop by drop, while stirring the mixture.

For this reaction, we didn't go for the Oppenauer oxidation, but because we like toxic substances, the oxidizer was the orange chromic acid solution.

1.1295 g of borneol were dissolved in 5 mL of ether in a conical flask which was placed on ice. Then, I added 6 mL of the oxidizing solution, drop by drop and gently mixing the reagents with the help of the magnetic stirrer.

Little by little the solution in the conical flask became darker, until, having finished to add the oxidizer, the flask was taken out of the ice bath so that the reaction could reach its end, at room temperature for approximately 15 minutes. Meanwhile, I turned on the magnetic stirrer every so often, at low speed.

The black-brownish colour results from the formation of reduced chromium byproducts. Predictably, this made impossible to visualize the bichromate (VI) ion (orange) being reduced to chromium (III) ion (green).


The first reaction scheme is little bit simplified: although the stoichiometry of the overall chemical equation is correct and really describe the whole process, this reaction is more intriguing and actually occurs in two elementary steps.

The former is the fast one: an alcohol (Borneol) and bichromate, at low pH, yield an ester (bornyl chromate). The latter is the slow, rate-determining one: here is where the 6 electrons are actually exchanged, as the oxidized ketone and the reduced chromium ion are formed.

10:30 am Everything is ready to kick off the extractions: the black solution contains our ketone (camphor) as well as the many byproducts.

I had prepared three beakers, to collect the three phases I would have separated, properly labelled them, placed the separatory funnel and checked the bottom valve. What could possibly go wrong?

However, when i poured the mixture in the funnel...DISASTER! Embarrassingly, I didn't close the valve completely so at least half of the liquid was spilled on the working table.

10:35 am Gentlemen, it's in times like this when men have to prove how tough they are!

As quickly as possible, I weighed another bit of borneol (1.1739 g), set to work as nothing had happened and cleaned the mess.

11:00 am Finally, I managed to pour my mixture in the sep funnel with no humiliating incidents. To make sure we didn't waste too much of our product, we were advised to help ourselves with some ether (10 mL) to wash the inner surface of the flask. Then, I added approximately 20 mL of water to begin to get rid of the chromium salts.

This first extraction was pretty tricky (especially at the beginning), because everything in the sep funnel looked the same and you could see the line separating the two solvents only when it approached the bottom tap.

At this point you might think: "Oh, there you are: you felt it was a race against time because you had that stupid accident!" Well, you wrong because just before my first extraction, other people were taken aback as, in their separation funnels a really weird phenomenon was taking place: ether was denser than the aqueous solution. They could appreciate this as they were extracting for the second or third time, so,  the organic phase wasn't brownish any more.

As it turned out, some idiot, probably on the day before, had filled a bottle of ether, with some dichloromethane too!

Honestly, I've no idea how they managed, but, as a result, I was quickly on the lead again. That gave the opportunity to work better: the aqueous phase was, in fact, put back in the funnel (the organic phase was collected in a separate beaker) and I carried out three extraction, each time with 20 mL of ether, whereas we were told we could have done it only twice.

Then, the organic phase was washed thrice with 20 mL of a saturated sodium bicarbonate 5% solution, to wash away, hopefully, all the chromium salts. A drawback could be caused by a third phase, between the aqueous and the organic ones, which consists of carbonates which have precipitated.
Happily, however, I have no such problem.

Some sodium sulphate was used to dry the solution of any traces of water in ether.

12:00 pm Things are going on nicely! I'm heading to the rotavap and I'm pretty confident that, given that this reaction is supposed to have high yields, my product will be also very pure.

A rotavap was used to boil away the solvent and concentrate the solvent. This was done at room temperature, since the solvent is ether and the melting point of camphor is relatively low (172-180°C).

 

 

 
Before I could assay the purity of the product, I weighed the flask and calculated the yield: the calculations gave me a massive 84.93% (0.9839 g).

12:30 pm Camphor has to be purified by sublimation: given that we don't have the proper instruments to perform this operation, we had to use our ingenuity.

I placed the round-bottom flask in a hot silicon oil bath and raised the temperature to 200°C: as the flask gets warm, I covered it with a normal funnel. So, as camphor sublimates, it reaches the colder funnel, where it crystallizes.

1:00 pm The purification was rather time consuming, but I managed to prepare the TLC with its mobile phase (hexane/ethyl acetate 4:1).

The melting point was measured in a bit of a hurry, so I can't give a precise number: what I have, instead, is a range (176-179°C), which is consistent with the data found in literature.

1:15 pm I literally threw the layer in the TLC chamber and got an astonishing result from the IR spectroscopy. The IR spectrum gave me further evidence that the product I had was really pure but, unfortunately, the purification wasn't over and I had to perform the TLC with some of the product still in the flask (hence, probably, less pure).

In fact, once the layer was washed with potassium permanganate and NaOH, it revealed a small amount of borneol was still present as impurity.

 

 

1:25 pm That borneol, however, came as no surprise and wasn't a big worry. The problem was that I had to clean an enormous amount of glassware in 5 minutes.

I can only say I'm sorry for the person who worked at that table in the afternoon... 

 

It's good to be back. After a short break, it's time to do some synthesis and, for my rentrée, I chose a simple reduction of a pretty common substance, camphor.

A very interesting feature of this reaction is that the result can be checked with a rather unusual instrument: your nose. That's why these molecules have different smells: not as different as chalk and cheese, but enough to be perceived by non-connoisseurs.

Camphor is extracted from the wood of camphor laurel (Cinnamomum Camphora), an Oriental tree, which belongs to the family of the Lauraceae. My favourite figure about this plant is the balsamic time: you can't begin to collect the precious oil, unless the tree is at least 40 years old. 40!

But it's not time to waste our time with flowers and agriculture, is it? It's time to crack camphor in order to yield some delicious isoborneol.

1.1542 g of camphor crystals were, easily, dissolved in 8 mL of methanol in a 50 mL conical flask. Soon after, I weighed approximately 0.6 g of sodium tetrahydridoborate, the reducing agent chosen for this reaction, while the alcoholic solution was placed on ice.

Sodium borohydride is a much more practical reagent than lithium aluminium hydride: although it's a weaker reducer, it's handier since, unlike lithium aluminium hydride, it doesn't react violently with water or methanol, so the reaction doesn't need to be carried out in ether or THF.
What's more, whereas the lithium salt is terrifyingly pyrophoric, our major concern with sodium borohydride was to avoid any contact with skin, given that this results in skin burns. This means that, annoyingly, I had to wear gloves...

Still, although we use a less strong and more selective reducing agent, it has to be highlighted that the reduction remains a pretty exothermic one: so, to make sure nothing might eventually go wrong, sodium borohydride was added little by little, constantly stirring the solution and with the flask on ice (keeping our reactants at O°C).

Once all the reducer had been dissolved, we heated the mixture under reflux for a couple of minutes at 40°C.

Isoborneol is the thermodynamically favourable product: yes, the hydride could attack the carbonyl on either side of the surface where the carbonyl lies, but the product with the equatorial hydroxyl group is more stable. Moreover, it's also the kinetic product.



At this point, as shown in the reaction scheme, what you end up with is an intermediate where boron is bound to isoborneol, instead of hydrogens.
To break this complex and yield the final products, we have to quench the reaction. So, 30 g of ice were thrown in the flask: not only does this cause a thermic shock (hence, quenching), but water, too, plays a key role, reacting with boron.

Because isoborneol is insoluble in water, once all the ice was melt, the suspension was filtered with a Büchner funnel. We can't, in fact, get rid of the water by just placing the flask in a stove, because this would melt everything.

Finally, the precipitate was dissolved in 20 mL of ether and the traces of water were completely removed using sodium sulphate.

The solution, then, was placed in a round-bottom flask and ether was evaporated in a rotavap, until the final product appeared, as a white solid on the whole innner surface of the flask.

To save time, we weighed the flask with its support (109.6420 g) and, therefore, calculated the weight of the yielded isoborneol, having previously determined how the empty flask weighed (108.8310 g).

Given that the yield had to be between 70% and 60%, I'm happy to report my colleague and I yielded 0.881 g of isoborneol, which means our yield is a wopping 69.34%!

Moreover, the product was outrageously pure, as either the melting point (213°C) and the IR spectrum revealed.

But this is just the beginning. Stay tuned! 

You spend a week working hard in the lab and look forward to the week-end. Finally, it comes and this should be the time when you enjoy yourself or study (or both, if you study molecular biology) but, unfortunately, some idiot has decided today the road in front of where you live must be resurfaced.

This may cause frustration and headache and you might need an Aspirin. Sure, you can buy it but what if you synthesized your own acetylsalicylic acid, instead of swallowing a tablet made by a hundred of underpaid Chinese workers?



To find out, we chose the industrial synthesis of this famous drug, that which starts from salicylic acid.



2.1506 g of anhydrous salicylic acid were placed in a 50 mL conical flask. Then, I added 3 mL of acetic anhydride and three drops of concentrated sulphuric acid to catalyze the esterification.

The mixture was heated by placing the flask in a water bath at 40°C for a couple of minutes. This would be particularly needed if a white resin (namely polymers of acetylsalicylic acid) precipitated.

I let the solution cool to room temperature and poured 30 mL of water: since Aspirin is insoluble in water, this will result in the precipitation of an amorphous product.

As it turned out, water wasn't enough, so, I put the flask on ice and scratched the inside with a glass rod to allow the product to precipitate.

Finally, it was time time to filter this first amount of Aspirin with a Büchner funnel.

To yield beautiful, needle-like, white crystals of acetylsalicylic acid, I performed a two-solvent recrystallization with ethanol and water: the former dissolves the aforementioned amorphous filtered precipitate, while the latter is the precipitating solvent.

I put this susance in two 25 mL test tubes, with approximately 10 mL of ethanol and then water: the limpid, alcoholic solution suddenly became white and opaque, as the precipitang solvent was added. The test tubes were warmed up in a water bath and the precipitate was dissolved once again.

We carefully stored the two solutions in a drawer to allow the crystals to form overnight.

On the next day, we filtered our crystals and weighed it. Apparently, the maximum yield for this reaction is 70%, but, sadly, we only yielded 1.3687 g (hence, 48.8%) of an outrageously pure product, as proved by its melting point (138°C) and IR spectrum.

 

I'd like to point out, nevertheless, that I got the highest yield in the lab...for what it matters.

 

Tea, what else? Xanthines are old friends of this blog. This time, however, I'm not going to talk about their properties: it's time to go deeper and see how much of these molecules is actually in your daily cup of tea.

The tea we chose for the extraction looked pretty old and wasn't placed in proper tea bags, either: a gram of tea from an old-fashioned bottle was ground in a mortar and 1.0511 g were accurately weighed.

Then, it was time to prepare a cup of tea. The pulverized leaves were put in a round-bottom flask, with 30 mL of water and nearby 0.5 g of calcium carbonate. The latter reagent is needed because xanthines (bases) tend to bind to tannins (acids): so, the carbonate displaces either caffeine and theophylline.

The mixture was heated to 120°C for 20 minutes. Meanwhile, I set to clean and regenerate the SM-2 column I would have utilized for the extraction. For this solid phase extraction, we selected a resin with the appropriate ratio of styrene to divinylbenzene (XAD-4) for the two lipophilic xanthines.

In a XAD-4 resin, in fact, the aromatic rings of either styrene and divinylbenzene are the perfect environment for lipophilic molecules such as caffeine, which can bind tightly, thanks to π-π interactions.



First of all, we poured 40 mL of ethanol and, then, cleaned the resin with 20 mL of water and 20 mL of bleach. From now on, the column will never get dry.

The procedure went on as follows: 40 mL of water, NaOH (1M) 20mL, water (40 mL), HCl (1M) 20 mL and then water until the pH of the column reached neutrality.

Once the tea was ready, we filtered it twice, in order to completely remove the carbonate.



The column was therefore loaded (twice) and the extraction of the alkaloids was underway, as we began to add the first amount (25 mL) of ethanol. We will carry out four extractions, with 25 mL each time.

Actually, the first 14 mL of (hopefully) only aqueous solution were collected separately in a couple of test tubes (which we were told to keep, just in case something had gone wrong), whereas the rest of the alcoholic solution was placed in a 100 mL volumetric flask.

Although (according to the protocol) the volume used for the four extractions should have been 100 mL, some more ethanol had to be added to reach the final volume.



In order to fill the column, I let the last quantity of ethanol go through the column: this increased the efficiency of the extraction and, in the end, this method proved to be more practical than doing the same operation with a pipette.

This solution was then diluted (1:10), to get a signal, from a spectrophotometric titration (at 276 nm), that could be compared with that of a 11.5 μg/mL solution of caffeine in ethanol,  our standard.

So, our 1:10 solution gave us an absorbance of 0.5632. This means the concentration of xanthines was 13.61 μg/mL; so, goin back to our sample, it  contained 1.30 % of xanthines.

Unfortunately, since the tea was rather old, nobody could find the original data from the manufacturer. Still, our professor said it was a plausible, albeit rather high, percentage.

Finally, to assay which xanthines we were measuring, we carried out a TLC (mobile phase: methylene chloride/methanol 9:1), once 25 mL of the first alcoholic solution had been concentrated to 2 mL at the rotavapor.