Skip to main content

Life in the Drug Labs

Chemicals, Shelves and Shelves of Them

My mention yesterday of the number of starting materials needed for drug synthesis prompted a reader outside the industry to ask just how many I might be talking about, and how these things are managed. I looked over the paper being discussed, just for an example, and for its three drug syntheses it needed a total of four commercial starting materials, twelve different reagents, and three solvents, and that doesn’t sound at all untypical.

What happens is a classic “long tail” distribution. There are reagents and solvents that are so widely used that they’re almost invariably kept in some sort of common storage area in a lab and/or found in smaller quantities at each bench. These are the potassium carbonates, sodium sulfates, ethyl acetates, and tetrahydrofurans – they come up constantly, and they need to be readily available. Past those are chemicals that are just a bit less universal, but that you can be sure of finding if you look in your own collection or the people next to you. In that category, I’d put things like the various palladium and platinum catalysts, lithium hydroxide, amide coupling reagents like EDCI and HATU, simple amines like morpholine or pyrrolidine, etc. If you run out, you know that someone’s got a bottle of the stuff around. They have to. It’s a med-chem department, and who ever heard of a med-chem department without a bottle of piperidine or some copper iodide around?

But after this, it’s long tail time. There are an awful lot of cyclic secondary amines once you get past those three I just mentioned, and at one time or another someone has probably wanted to put one of them on their scaffold. So these things are stored in the stockroom, an ever-expanding collection of commercial reagents and in-house intermediates. Every good-sized chemistry department has to devote effort, staffing, and resources to the problem of storing and keeping track of these things. I’ve seen a lot of different systems used over the years, and it’s safe to say that no perfect one exists.

One thing that varies is how many reagents people are allowed to keep at their own bench or hood. Some places are OK with people having what are nearly mini-stockrooms of their own (“Go see Pete; he’s got a big collection of stuff like this”), while other departments actively discourage this sort of thing and practice “reagent decluttering” as a rule. No matter what, though, every bottle has to be marked and noticed, usually with some sort of bar-code system. But quis barcodet ipsos barcodes? Someone has to keep track of how the reagents move from the labs to the stockroom and back again. Even worse, the chemists are constantly going to be borrowing stockroom reagents from each other and not returning them to the owners of record, because that’s what we do. In my own experience, depending on the chemists themselves to keep these inventory systems up to date is a doomed endeavour. We have other stuff going on, and we’re always going to just return that reagent as soon as we’re done with it, or bar-code those bottles that came in just as soon as we’re done with that other thing. Run things like that for a few months, and no one knows where anything is at all.

You need people who do nothing else but keep track of the compounds, a staff to round these things up and keep order back in the stockroom, and you’ll have to make sure that the chemists are kept out of there or you can kiss that order goodbye. What you’ll usually see is some kind of ordering and lab delivery system to each bench, and on the other end, bin or designated area for reagent return, which is sometimes divided into “Back to the Stockroom” and “Empty/Dispose” categories. All this costs money, but if you don’t pay it, you’ll pay in the less visible but even more expensive currency of people wasting time wandering around looking for reagents, or people just not setting up particular reactions because they can’t find the darn coupling partner and the heck with it. Even the better-managed shops need to call a halt occasionally for an Inventory Day or two, where everyone’s bench, hood, cabinets and fridges are scanned to reset all the bottle locations and ownerships.

Another aspect of inventory management is dealing with duplicate bottles and clearing out things that are no longer useful. A good inventory system will cut down on the duplication, because people will be able to see that there are already supplies of Reagent X in house (although some people will always order a fresh bottle just for themselves). That fresh-bottle behavior is particularly acute, naturally, for those reagents that are known to deteriorate. A jar of sodium bicarbonate from 1927 is just as good as it ever was, but a bottle of aniline from last year will already have turned red or even black by now. It was probably at least yellow when it came in the door – freshly purified aniline is clear, but I’ve only seen a couple of samples of clear aniline in my life, because it reacts with atmospheric oxygen and it’s all downhill from there. (If you go back to the really old literature, you can see substantial confusion about just what aniline’s boiling and freezing points really were, because everyone was working with sample of different purity).

Oxygen and ambient humidity are what wear on most reagents that are capable of wear (some amines will pick up carbon dioxide as well). The beloved “palladium tetrakis” catalyst (that we all use even though there are better catalysts out there for any given palladium coupling if you look) is another example. A fresh bottle is a beautiful light, clear lemon yellow, but over time the bloom fades, a bit of orange creeps in that will sing in deeper and sadder tones of red and brown if you give it long enough, and if you’re chucking brown tetrakis into a reaction and expecting it to work, you are a lazy hack indeed. Generally, it’s the chemists out on the bench who decide that a given bottle of reagent has lost its zing, and are expected to dispose of it and put it the empty bottle back in the bin to be removed from the system. Personally, I have tended to order the stockroom bottles that have newer-looking barcodes on them in the inventory software, and I’m sure I’m not the only person who does it, either.

Another feature of a long-running stockroom is the number of reagents in it that were once articles of commerce but are so no longer. These things (as long as they haven’t gone bad) are actually quite valuable, as you’d imagine. I also enjoy the sight of older label styles (and apparently I’m not alone). Once in a while I’ll see the “good old Aldrich” labels from the 1980s (and probably before) when I was first starting to work in the lab. (A slightly modified version is here!) In fact, you can sometimes even see labels of whole supply companies that have disappeared – for years, Lancaster Synthesis bottles would appear in my stockroom orders even though that name was no more to be found in the real world.

So that’s a long answer to the question of “How many chemicals do you people have around?” The answer is thousands, of all degrees of purity, age, popularity, and utility, and keeping track of them is a full-time job!

53 comments on “Chemicals, Shelves and Shelves of Them”

  1. Cato says:

    No need to ever use brown tetrakis–a simple wash with cold methanol will put it right back to its canary yellow glory.

  2. Mad Chemist says:

    As an academic organometallic chemist (for now, I want out as soon as I graduate), we use a lot of acetonitrile and diethyl ether. I rarely use ethyl acetate, other than for chromatography. We also have multiple versions of most of the first row transition metals, and many of the second and third row metals. One of our problems is gauging the age/quality of whatever silver salts we want to use, due to photoreduction. We also work with organic azides, and that’s another thing that can generally be judged by color.

  3. SP says:

    We were about to acquire an RFID system- not quite as simple as the Amazon Go setup where you walk just out the door and it knows what you have, but you collect your bottles from the stockroom in a tray, swipe your badge at the computer at the door, put the tray on the pad, and the tray contents are assigned to you. Then you walk to wherever you’re going to work (usually your bench) and repeat the process to log them there. On the purchasing side, you check a box when you order a chemical you want logged, on arrival the shipping & receiving group applies the RFID label and logs it into the stock room. On disposal the waste company logs it out.
    Management decided that was too extravagant- $1 to $2 per barcode!- and canceled the project and went back to handheld barcode scanners.

    1. BrownTetrakis says:

      $1-$2 per barcode??? That seems high… if it’s cheap enough for Walmart, it should be cheap enough for us chemists, right?

      1. ata says:

        Walmart buys a lot more RFIDs than a pharma company…bulk discounts. What we really need are augmented reality glasses that automatically read whatever the chemist grabs from the shelf and then tracks it (reagent, lot number, barcode, etc..), matching it to a real time inventory and automatic ordering system.

      2. matt says:

        Walmart actually backed away from the RFID system, anyway. They use plain old paper barcodes. There were severe issues with disambiguating multiple echoes along with strongly variable signal attenuation. I think they might have stuck with RFID for clothing, but not much else. I suspect the Amazon store RFID checkout will go the way of its drone package delivery, and Google’s blimp/balloon based Internet hosting, and Popular Mechanics/Science’s flying cars for the family: any actual implementation will prove the impracticality of the idea for general use.

    2. Derek Lowe says:

      I think that eventually RFID is the way to go, but you’re right – it really is too expensive now. But I can imagine having the reagent bottle track itself through the labs as you walk around with it – now that’s my idea of the future. As long as they don’t stick one on my forehead.

      1. SP says:

        The price was fairly high due to scale and that we needed somewhat more robust barcodes as there was a question of transmission through various solvents etc. depending on where the label was. But even so, we estimated ~$15k to tag bottles initially, then a few thousand more per year, which would have put us close to the $1 per label range. We’re not a terribly large operation but even that cost would not have been very noticeable in the budget- the software and hardware were maybe another $30-40k up front. It really wasn’t much compared to salaries or instrument service contracts.

        1. tangent says:

          Wow, just guessing from afar, but it seems like $1 might pay for itself *awfully* quickly — I mean what’s the $/hour rate that they use for a chemist, and how hard is it for an RFID tag to save a few minutes just out of routine handling overhead, much less in slightly less often of having to look around for the bottle of whatever-it-is.

          Yeah yeah I know that’s not how spending decisions actually happen.

          1. SP says:

            Hey, would you like to be my manager?

      2. JIA says:

        Derek, they won’t stick it on your forehead — they will inject it into your forearm! See news link in handle.

      3. Jim Mowreader says:

        We’re working on the assumption that whatever’s in the bottle won’t block the RF signal from reaching the transponder on the bottle. I haven’t tried it (having no access to an RFID system), but I suspect the more expensive a reagent is, the more likely it is to act as an RF shield.

    3. David says:

      Walmart probably makes their own and/or uses billions or trillions a year anyway. From memory, I remember hearing that they cost the store somewhere in the region of 10-25 cents when I worked there.

    4. NHR_GUY says:

      You can have all the fancy schmancy technology you want for keeping track of inventory but I have always said this “an inventory system is only as good as the people who use it”.

  4. anon says:

    “coupling partner and the heck with it”
    I see what you did there. 🙂

  5. Bagger Vance says:

    Does anyone have a management system they recommend for a very small lab?

    1. sdkj says:


      1. Etienne D says:

        😀 nice one!

    2. Peter B says:

      Look into FindMolecule for chemicals and Quartzy for other consumables. Both free.

    3. Andrew S says:

      You could try ChemInventory. We use it in our lab (academic) and it does everything we want it to.

  6. Teamplayer says:

    1. If you order a chemical and “hide” it in your area, remember, it’s not your’s, it belongs to your company.
    2. If I take a chemical from your area, be assured, I will let you know especially if you are not around at that moment. I will also order another one for you to replace yours if needed. I expect others to do the same for me.
    3. Numbers 1-2 above can also be said for uncommon to almost common equipment and glassware.
    4. Minimize duplication and waste. Common areas and common storage systems do actually work rather than you squirreling away “your stuff”. Don’t be old-school.

    1. neo says:

      Dude, even if I’m the third string quarterback for the Minnesota Vikings, once the team buys me a jock strap, it’s MY jock strap, and NO, you cannot wear it. I’ll happily share my towel, or my jersey. There are some things I’ll share and a few things I won’t. You’re more than welcome to my ethyl acetate, but hands off my opened (and carefully resealed) bottle of butyl lithium.

      1. Mister B. says:

        Same for my trimethylaluminium solution !

        Lack of confidence with others or lack of training from others ? 🙂

      2. picky says:

        I squirrel stuff away because everybody doesn’t treat their chemicals with the same respect. Anhydrous solvents and air-sensitive reagents suffer from tragedy of the commons. If you mess them up, you’re actively hurting my ability to perform my job.

        1. Teamplayer says:

          Neo, you’re not a team player. Play rugby then you’ll understand

          Neo, Picky and Mister B you’ve all just admitted you cannot trust your coworkers (or even their results now). Appropriate training or use this other great communication device called the mouth.

          1. picky says:

            It’s usually not trusting the one weird coworker, the one who does not take to training and doesn’t respond to said communication device. Every lab I’ve ever worked in has one of these.

            You’re more than welcome to the stuff I have squirreled away. But it’s hidden so Igor doesn’t get to it.

          2. Hap says:

            People have known not to drive drunk for a very long time now, because it’s illegal, potential costly and homicidal, and dumb. Messages not to do it are broadcast on every medium, pretty much. Yet about ten thousand people a year get killed doing it (or killed by others doing it) where I live.

            You can’t cure stupid (or a%&**%). You also generally can’t cure selfishness. If your plan depends on these being cured, it’s not going to work.

          3. tangent says:

            The question is, during the hiring interview can a clever line of discussion get somebody to admit they don’t treat their anhydrous reagents with respect. Interviewing technique tips for how to approach it obliquely and like maybe you do it yourself, just between us chickens?

          4. anon says:

            Appropriate training always starts with “don’t trust anyone”

    2. Derek Lowe says:

      Common chemical areas make a lot of sense, but if someone wants to have their own well-maintained Sure-Seal bottles of air-sensitive reagents or carefully stored catalysts, I say let ’em have them. The amount of money involved is not something to worry about, in most cases.

      I would also say that putting a bottle in your hood is not hiding it. I assume that anything in someone’s hood is being actively used, and that anything in their cabinet is something they actually need or care about. I really don’t want to take such things when the person isn’t around. I once ordered a fresh 800 mL bottle of borane/THF, all of which was needed for the first step of a big prep. As I discovered the next morning, someone came and took it from the fridge the very afternoon it arrived and helped themselves, and there went the timetable for that.

      I dislike both extremes – every chemist being their own stockroom doesn’t work, but I have trouble with the happy-commune model as well.

  7. David says:

    A scanner at each station to be utilised every morning, or a designated person to wander about scanning once a day would go a long way (assuming you don’t want to spring for a scanner for every station). Complete with 4 buttons indicating quarter through full status, if you want to be fancy?
    Assuming expiration or “check if it’s still good” dates can be put in the system on acquisition or synthesis if it’s in-house, the people responsible, or each person at their station could spend a few minutes a week checking on those reagents whose use/check by dates came up this week?
    Used containers could be send to the inventory folks or recycling/trash/cleaning and reuse, together with scanning out of inventory.

    My naivety thinks it shouldn’t be so hard..

  8. UudonRock says:

    I’ve developed a few inventory systems in the past, mainly for academia but also for industry. Out of necessity these were generally on the cheap and used what was at hand since upkeep on a “proper” LIS is not in the standard budget of many small to mid-sized colleges. I implemented a weight count system based on QR codes that had a double purpose for one small college. An optical scanner attached to a scale in the stockroom would tare the reagent as it was signed out and re-tare as it was brought back (didn’t work too well for those barrels of acetone though…). Each user had a code to sign in to the count system as they took chemicals out of the inventory. The system manager could look at the list and see what was signed out, by whom and what needed to be ordered. The codes had a link to a second database that would provide SDS (MSDS at the time…) through a phone app. The manager had to generate these codes and allocate them to the proper reagent and MSDS in their respective databases. The system worked extraordinarily well. But. As soon as one person stopped using it the whole thing fell apart. There is no perfect solution to the self-problem, and the management thereof. It only takes one jerk to ruin it for everyone. We all work with at least one of those.

    1. ScientistSailor says:

      That’s right, as soon as you make something idiot-proof, they go and build a better idiot…

      1. eyesoars says:

        There’s very little that’s malicious-asshole proof. 🙁

        1. UudonRock says:

          Maybe someone should apply for a grant to figure out who keeps making all these idiots and assholes so they can be asked to stop or be forced into politics where they belong.

          1. Hap says:

            Problem is, they don’t want to stay in politics – if you elect them, they write the rules so that it’s easier for them to spread and get to the money, kind of like flesh-eating bacteria breaking the connective tissue up so they can get to the really good stuff. They’re not going to go away, but mainlining them into the bloodstream won’t help. Maybe there’s a Fletcher Memorial Home they can stay in.

  9. Steward says:

    For our chemical storeroom, I have noticed they started to put new barcodes on old reagent bottles but they only mark returned stuff as half-full in our system. I make a conscious effort to use what’s available, barring any long term stability issues, and sometimes by doing so I can properly dispose of that bottle that has only a dust coating or evaporated over time. I learned to order at least two bottles for that reason.

    Another thing i noticed are the endless supplies of Buchwald ligands. Most of our organization uses the precatalyst versions for ease of use now for Med Chem work, unless they have failed to create such a precatalyst.

  10. Anon rabbit says:

    There’s always the aspect of reporting the compounds to the local authorities (fire, environemental, …), keeping track of MSDS (much better now), and with our system, we have reports to clear our stock of peroxide potential compounds (ie. degration product are peroxides).

  11. I visited Edison’s lab and workshop in West Orange, NJ last year. In his chem lab they still had hundreds of bottles of chemicals just sitting where they left them when the place closed in 1931 after Edison died. I asked the Park Ranger (this is actually a natural park) about any of the reagents decomposing over time, and he said that at some point chemists have gone through it all to remove the dangerous ones. Still, pretty cool to see it exactly as it looked back in the day.

    1. oldname says:

      I’ll bet that the chemists doing that assessment were on first name terms with the bomb squad.

  12. Obviously a joke says:

    The obvious solution is to put all the chemicals on the blockchain, which is what my new startup is planning to do. ICO to come.

  13. Mike says:

    Life is a constant battle against entropy. Soldier on everyone!

  14. Chrispy says:

    You chemistry folks have to help me out here: my wife contends (and is supported by all kinds of internet sites) that baking soda goes bad. Not that it gets an off-taste, mind you, that it stops leavening. I have trouble believing that it becomes a box of sodium hydroxide without becoming a solid block in the process. If it is a fluffy, white, powder, isn’t it still baking soda? (Baking powder is more complex — I’m talking straight NaHCO3.)

    1. Derek Lowe says:

      Baking powder definitely goes off, but that’s because it’s a mixture of sodium bicarbonate and some sort of acidic salt (a phosphate or sulfate). Those will react if the mixture picks up moisture from the air, and the result is a less effective mixture.

      But baking soda itself, sodium bicarbonate? Nope. It is what it is. It’s not hygroscopic, and doesn’t react with water, anyway. The only common thing (besides an acid) that will cause it to break down is heat, and you have to be up to oven temperatures to get appreciable rates of decomposition. (I do wonder about a box that’s been stored near the Persian Gulf for a hundred years or so, but I don’t think that’s what we’re talking about).

      Baking soda does not go bad. I know that there are hordes of internet sites that say otherwise. They’re wrong.

  15. cynical1 says:

    From my experience you should treat every chemical as if it was the most precious chemical you just ordered. Parafilm it even if it probably doesn’t need it (within reason). Put it under nitrogen or argon before you do it. Mark when you used it (like for Grignards and organolithiums) and even the amount on the label of the bottle.

    And then assume that every other chemist is a selfish a**hole who would never do the same when you get a reagent from another lab. And look at the chemist who had it checked out before you. If he/she ordered it in the first place and you trust them, then you can probably use it without worrying too much.

    1. Chester says:

      The only thing that pisses me off more than no parafilm is bad parafilm. We all know that one idiot who wraps 15 layers of parafilm around the lid of the bottle yet doesn’t press it down and leaves all sorts of gaps for air to get in. What even is the point of doing that…

      1. Derek Lowe says:

        I have to agree – nonstretched mushed-on parafilm is the mark of a cretin.

        1. Dave says:

          Of course chemists love to layer parafilm on their reagent bottles… but has there ever been a study that shows that it actually accomplishes anything meaningful?

  16. tangent says:

    quis barcodet ipsos barcodes?

    Derek, I don’t want you to think no one noticed that little specimen. Or else you might escalate.

    (Really, cute 3rd nom. pl. morphology.)

  17. jwskud says:

    One possible useful tip to add to all of this – whenever I use a chemical, I use a Sharpie or attach a small adhesive label to the lid of the bottle with the name (EDC, HOAT, nBuLi, etc.) or drawn structure of the reagent/building block. This labeling system makes it much easier to find chemicals using a “top-down” approach, instead of having to pull each bottle up to look at the labels. This is especially useful for finding things in loaded bins, loaded desiccators, loaded freezers, etc. I can find things in a few seconds that would otherwise take many minutes.
    I once asked an Aldrich representative if they could print the name/structure on the lid of their chemicals and they said it technically difficult, so for now, I rely on this method to find things quickly.

  18. heteromeles says:

    I was just thinking that instead of automating reactions, you need Marvin the Automated Stocking System (bad Hitchhiker’s Guide joke). More seriously, it might be somewhat easier to automate QA/QC than to automate synthesis, and simply analyzing chemical use and ordering habits might be one (at least touted) as one way to cut down on expenditures, by ordering what chemists habitually use instead of what they think they need. This, of course, runs into the small problem of exploration being unpredictable, but I suspect that machine learning might be a better way to clue into habits, anticipate orders, checking stock quality, and the like.

  19. Papa Francesco says:

    You must treat compounds as you do all da peoples- With Love and Respect!
    If you neglect dem, they will decompose and all go away!

    Love All Molecules like dey your own Brothers and SIsters!

  20. Scott says:

    Some of us are quite happy that various amines (like monoethyl amine) absorb CO2 from the air. They keep the air breathable, if distinctively aromatic, in a submarine!

    (Though I am concerned that every once in a while I manage to find something and get a noseful of ‘Eau du Boat’, since I haven’t been to sea since 2004!)

Comments are closed.