SMS (archive)

In Contrast to Genesis and Excelsior this project was not born from the need to run several hundred reactions in a month, but rather from the sheer will to join the automation movement in chemistry by automating a time consuming task of chemistry.

While standard operations like column chromatography have automated solutions readily available, there are no such options when it comes to syntheses themselves, because this is what a chemist is for, right? No! A Ph.D. student spent about 5-7 years on his/her education. When those people have to redo an experiment that was already conducted multiple times, it is a waste of human resources, which is the worst kind of waste possible. Still, when it comes to performing basic organic reactions, everything has to be done manually. If you are lucky you have an undergraduate student, who can use this as a chance to learn something new, but what to do when there are no students available due to a… lets say… global pandemic. You dont want to waste your time on a reaction that you did 5 times already.

So lets start building a synthesis robot that solves all of these problems! First, find a nice place for your robot to live.

What now? I know! Reading literature is always a good thing to start with.

Literature

Of course there are automated systems for syntheses, but generally they are so expensive that only the industry is willing to buy it. The most recent version of an automated batch synthesis platform from -not for- the academia was published 2019 by Leeroy Cronin, who built the so called Chemputer (which is also the name of the general concept).

It looks really cool and practical at first, but when you examine how everything is built, pretty soon you will realize that about 50% of all parts are self crafted (3D printed / special work from the internal workshop) which makes it complicated to reconstruct. Furthermore, this whole system is marketed in a spin-off company from Cronin, where it is sold again for a price no research group in academia wants to spend.

Construction

So what we want to build is a platform with as little money and self-crafted parts as possible. Alright, what do we got so far?

Damn! Still nothing. Lets get some devices noone is using anyway. Get something to stir, a valve and a pump – that should cover the basic functionality. Now make them communicate with each other. How hard can that be?

Turns out pretty darn hard when you do not know a thing about serial protocols and all that stuff. But finally, after 2 months those devices can be controlled via PC. Do not get this the wrong way: The main achievments here are not the working devices but the knowledge that allows us to control every piece of lab equipment with a half-way decent documentation!
After planning the system, we can start collecting more valves and some glassware, plug it together and voilà: a great-looking setup.


Now that looks pretty cool! The downside is that right now it can do absolutly nothing. Everything can only be controlled individually from the PC. If you need to click everything in sequence for it to happen, you did not automate, but mechanized it. So now we need the logic of a synthesis. Everything to the smallest. Possible. Step.

Logic

Say you need 150 mL of a chemical. So tell the pump “move 150 mL” and then it shou…. *brrrr* you failed. Nothing happened. The pump has a moving range of 3000 steps which scale to the syringe size. In this case just run 3000 steps up and down until you reach… *brrrrr* you failed. Nothing happened again. Switch the valves in between each moving cycle or you pump just back and forth. Ah yes, now something comes through the valves into the reaction flask! Now I can rule over all of… *system goes brrrrr* you failed. Everything blew up this time. You used a chemical that was not inert towards one of your system’s components, or you had a little bit of solid in the system, or its just not your day and your devices hate you.
Authors remark: Originally this paragraph was a lot longer to show more of the problems possible, but I will just cut it here.

So lets say you implemented all functions and debugged until everything worked in unison. Splendid, now everybody is in awe about your creation and wants to use it for their synthesis. That is not the case for two reasons:
1) The first reason is that right now you are the only person that can use the system. Noone else has a clue what is does or how. Go write an interface that normal people can use. The easier the better. Here a simple csv file with human readable commands was chosen. Then go present it in an understandable way and give examples of everything.
After a lot of work, presentations and time went into all of this, still noone really wants to use it. Why?

The other reason

Authors remark: To not step on some toes (or feet) I will only quote already puplished papers and articles.

When confronted with new technologies most of the current working chemists are sceptical to say the least. Generally claiming that fully automated organic syntheses are an insult or impossible to realize.¹ Still when technoglogies compare to humans the silicon based candidate wins everytime in all categories (retrosynthesis, optimizing reaction conditions, running syntheses batch and flow) and if not now then within the next few years, since computers are getting better a lot faster than humans.²
Despite these rapid improvements, the more traditional groups still rely on a “fraction collector that has a name, takes a lunch break, and collects a salary”¹

Most chemists are not trained in computer science, and might feel scared about such aspects and perhaps a bit threatened
– Mimi Hii, Imperial College London³

Some people view chemistry as an art. I think this might be a barrier to people embracing automation
– Rachel Grainger, Astex Pharmaceuticals³

Summary

Over the next few years we will see an increasing amount of technology in all areas of life. Alongside those discoveries all scientific fields will adapt at least some of the improvements. This will lead to more technology based study programms, which in the long run will lead to discoveries people 20 years ago could not even dream about.
Until then do not feel threatened by anything that is not your specific field of expertise, but learn whatever you can and use it.

Contact

If you want more infos towards the technical part:
Code: https://github.com/Kaffejunge/Master_Project
Full master thesis/manual: Write me an informal mail to finn.bork@wegard.de

Sources

  1. Automated Chemistry: A Vision
  2. Organic Synthesis: March of the Machines
  3. A brave new world of robot chemists and ‘synthesiser farms’ awaits