The Sheri Sangji Case

Many readers know that research assistant Sheri Sangji died from burns sustained in a laboratory fire in the lab of UCLA professor Patrick Harran. Harran and university Regents are up on felony charges for their part in the incident. I understand that the charges are based on occupational health and safety violations related to the incident.

[The excellent blog Chemjobber has been following this story.  I might add that this blog should be put on your Favorites list if it isn’t already there. The author puts a lot of work into it and it shows.]

Sangji was transferring t-butyllithium when her plastic syringe came apart and a quantity of the pyrophoric solution was splashed on her and ignited. She sustained fatal burns when her clothing caught fire and she died 18 days later.

Syringe techniques are common and the use of plastic syringes in such transfers of lithium alkyls is not unusual or automatically over-dangerous. However, some syringes have what is called a Luer tip where a syringe needle is attached solely by friction.

Another design has a Luer lock where the needle is affixed with a twist of the needle into a friction lock.  The former design, with the tubular tip and no locking mechanism is prone to disconnection under tension and on withdrawl of the needle from the septum on a pressurized bottle, the needle is likely to squirt bottle contents onto the worker. The Luer lock largely prevents this type of accident.

Another failure mode is when the plunger is inadvertantly withdrawn completely from the barrel of the syringe. Minimally, this would release the contents from the barrel, possibly on the operator. If the plunger is pulled completely out while the needle is still in a pressurized bottle, a fountain of liquid may discharge, possibly on the operator.

Syringe plungers with a rubber tip are prone to swelling in organic solvents and may become difficult to move during a single use. If the plunger is pulled with great force, it might release suddenly causing it to come out of the barrel along with the contents.

Other syringes have plungers that provide a seal by plastic-on-plastic pressure. The seal depends on the elasticity of the barrel to accomodate the slightly oversized plunger. These syringes do not come with Luer locks and as such, are not forgiving of less than skillful use.

I do not know exactly what technique Sangji was using. Aldrich distributes literature on the use of a cannula in the transfer of air sensitive liquids. That is fine, but if you want 0.1 to 60 mL of RLi, a syringe is the most expeditious method for delivering a precise aliquot in my opinion.

Experimentalists are often stricken with a cowboy mentality. If you have never had a serious incident with a material, it is easy to get a bit cavalier. But handling metal alkyls is a lot like handling rattle snakes- you have to be careful every single time.

A subsequent post offers suggestions on due diligence for ressearch professors.

About gaussling

Gaussling is a senior scientist in the chemical business. He occasionally breaks glassware and has been known to generate new forms of hazmats. Gaussling also digs aerospace, geology, and community theatre. View all posts by gaussling

18 responses to “The Sheri Sangji Case

  • Coffee Lover

    Hmm, I’m reading this and trying to see your point, Gauss. Please speak directly – I am very slow to learn.

    Are you suggesting there is an opportunity here to make a better syringe? Do you remember the syringe pumps – how they leaked overnight? It was always a crap shoot as to whether your reaction would be successful. The rubber bands and springs to hold the plunger in place!!! What a life.

    As a faculty member at a PUI who has a bottle of said tert-butyl lithium in the fridge, I make it clear that if the bottle is going to be used I have to be present, because I’m the one who is going to have to talk to the parents afterwards. But, you know – I just saw today where one of my older students is off running reactions without talking to me. I think I am getting nervous.

    • gaussling

      You mean, you didn’t like sopping up puddles of ethyl diazoacetate? I remember when Onus wanted me to react diazomethane with phosgene. I told him I wouldn’t do it. Thinking back, I should have done it.

    • gaussling

      OK. Good point about no point. I guess I don’t think it is unreasonable for grad students and up to use a syringe for t-BuLi transfers. [In grad school most of us at least once took a syringe load of tBuLi and squirted it through the air into some kind of containment- a large beaker with ice in my case. It made a lovely parabolic arc of flaming liquid. Kinda stupid, I guess.]

      Some syringes are safer than others. Using a non-Leuer lock syringe is a choice that offers more risk that otherwise. Glass syringes have a bad habit of seizing with air sensitive solutions. Plastic without a rubber plunger is the best choice.

      That being said, The professor must realize that his/her underlings are handling reactive materials and act in a responsible manner to provide oversight accordingly. But eventually you have to trust the chemist to not hurt themself. Management by walking around is useful.

      The issue of personal protective equipment (PPE) arises here. Synthetic clothing is a problem. It burns and melts on the skin. People working with flammable substances should be in fire retardant clothing. A lab must have a culture of donning PPE and that starts at the top.

      If professor Harran is like most professors, his career has been spent in academics where OSHA requirements do not penetrate. Shallow planning and training for the handling of metal alkyls is perhaps the norm. It is normal for profs to let more experienced students and post-docs do the actual training.

      Every lab should have a copy of Shriver’s “The Manipulation of Air Sensitive Compunds”.

      The student that is running reactions without advising you is exhibiting a form of initiative. You just need to nudge him/her back into the fold by saying that you’re curious to see his/her great technique.

  • Old Chem Prof

    Bravo. Thank you for your commentary. It is much needed.

  • agogmagog

    ‘Some syringes are safer than others’ – often found out the hard way. I preferred the 10ml glass syringe (big hands helped). Others in the lab liked the 5ml plastic with luer lock and no rubber seal. Rubber seals = trouble.

    Here in industry I prefer the chemflex cannula system from Aldrich. We scale our process to the pack size so there are no left overs.

    I loved the fact the at university I was allowed to use any reagent I saw fit (though the line was drawn at phosgene, because triphosgene was avaialble). The more dangerous ones, Lithal, t-BuLi, CH2N2, KCN etc were used, for the first few times, under the supervision of my boss.

    We were given the most freedom when we were the most inexperienced (and most enthusiastic)and it would sadden me greatly if this freedom to explore dangerous ground was curtailed.

  • paul

    In my experience with pyrophorics (t-buli, neat TMAL, DEZ etc…) I have found that the PP syringes that students are accustomed to are unacceptable, there is too much variation in the barrel/plunger seal and the luers have inconsistent friction. The glass barrel / plunger ones appear on the surface to be a better choice, but getting a stuck plunger due to air/water reactivity is common. The best bet are the glass barrel/metal plunger syringes with the PTFE seals (and the luer lock tips, preferably with a luer valve). These offer 2-3 PTFE seal rings and enough barrel friction so you have to think when you use the syringe – it is not easy to use one handed!

    I also feel that proper Schlenk techniques are more of an art that a skill. Often, the student learns on a particular set up and then cannot reason through how to accomplish a transfer on a different vac/Ar line system.

    • gaussling

      You’re right about the glass barrel/metal plunger syringes with the PTFE seals. I forgot to mention them. Undergrads with pyrophoric materials is a bad combination. A person should have accumulated a few years of lab experience before they are allowed to use t-BuLi without direct supervision. It takes some lab savvy and a bit of mechanical aptitude.

  • Chemjobber

    Thanks for the kind words, gauss.

    I hope that the Sangji case will eliminate the cowboy mentality w/r/t the alkyllithiums — I think of it every time I use a syringe.

  • Uncle Al

    Needle + nasty = bad when withdrawn. Take an inch of 7mm or smaller tubing, fire polish the ends, and insert a tight tiny septum in each end after argon flush. If paranoid, wire the septa. Needle goes through both septa, the distal one snugged to the bottle’s septum. Load syringe (argon ullage. Push in a gas volume equal to the liquid volume to be withdrawn). Pull needle (snugged to syringe!) into the chamber. If it leaks or sucks a bit, so what? Snug distal septum to the reaction, push needle through and inject. Withdraw needle into chamber. Your butt is covered in all directions. Neat Me3Al is a whole different monster. Know when to be in a glove box.

    Science 335(6068) 544 (2012) “Ready, Steady, Compete”

    “Without affirmative action, a competitor won the tournament if his/her performance exceeded that of at least four of the five other group members. [Objective qualification] With affirmative action, the two winners were the best-performing woman and the highest performer of the five remaining participants.”

    NOW, THE FUN PART! “In the latter case the entry rate of high-performance women into the competition rose, whereas that of men fell.”

    Imagine that – being told first place is awarded not earned discourages those who do the earning. Tell the fried Affirmative Action female that she was awarded an “A” in the course.

  • Morris

    I thought that I read in one of the reports or writeups that a ~1L beaker of hexane was spilled over once the BuLi lit up. Seems to me that this would have been a major contributor to the terrible injuries; it was the majority of the fuel- while the BuLi would have been only an ignition source.

    I make it a point to remind new operators that these flammable liquids that we see everyday are just as dangerous when spilled as gasoline… often much more so.

    • gaussling

      I think your reminders about the flammability are right on. I have seen several burned out hoods and it is always evident to me that it is best to keep the amount of flammables to a minimum in your workspace. The fires I have investigated have damage that scales with the amount of liquid flammables available.

  • Jyllian Kemsley

    > I do not know exactly what technique Sangji was using.

    Sangji was using a plastic syringe with a 1.5″ needle to do three transfers of 50-60 mL tBuLi (total amount was to be 159.5 mL). More at:

    Syringe remains shown here:

  • Friday chemical safety round-up | The Safety Zone

    […] discussed syringe use and academic lab due diligence […]

  • A. Anderson

    There are two types of syringe commonly in use: plastic and glass. Both have their advantages and drawbacks. With plastic, there is swelling. This can become so bad that the piston cannot be moved. However, for aqueous solutions, that is not a problem. Glass has the obvious problem: it can break. While plastic syringes can take a lot of force, applying force to glass is always a very bad idea. Let’s assume that the mechanical problem is that some force has to be applied to the piston. This may be due to swelling, blockage of the needle, or crystallization of material between the inside wall of the cylinder and the outside wall of the syringe piston. When movement is restored because of the application of force, the piston may now be pulled in an uncontrollable manner and “pop” from the syringe cylinder. The answer to this problem is to limit the travel of the syringe piston so that it cannot be removed from the syringe cylinder. I have found two solutions, one for glass syringes and one for plastic syringes. In the case of glass syringes, this involves carefully wiring a loop of single stranded 24 gauge steel galvanized wire onto the outside of the glass cylinder just below the glass cylinder flange. Then do the same thing with another wire to the glass piston. Many glass pistons are “necked-down” just below the piston flange that is pressed on to move the piston, so that is pretty easy. If the piston is not “necked-down”, then a slightly more elaborate wiring job is needed to cage the top piston flange. Having done that, the two wires are twisted together and limited in length so as to prevent excess travel of the syringe piston. The wire that I used is available in most hardware stores and is found where picture hanging supplies are sold in the store. The wire is quite flexible and strong and will not hinder loading or ejection of liquids from the syringe. It is not so stiff as to cause glass breakage when wired to the glass cylinder. The same solution can be applied to plastic syringes. However, in the case of plastic pistons, which are usually not “necked-down” below the piston flange, a small 1/16 inch hole is drilled through the top flange to secure the top wire to the piston. Then the two wires are again twisted together to limit travel of the piston. The wires can easily be cut and removed for syringe cleaning.
    Disclaimer: I cannot take responsibility for the use of the above mentioned technique to limit syringe piston travel. Some people lack the mechanical ability to fabricate such a device, act in haste, or have no plan for what to do in an emergency. Some people seem to have a divinely inspired ability to screw up even simple directions. It the responsibility of senior researchers to make sure the syringe is wired properly to limit piston travel, the wire is free of corrosion, etc., and that inexperienced workers understand the benefits and limitations of the device.

    • gaussling

      Thanks for the great suggestion! Wiring the plunger so that it cannot pull out is a good layer of protection and easy to do.

      Now that I have seen Jyllian Kemsley’s citation above regarding the amount of liquid that was being transferred, I would offer that the transfer should have been done by cannula using the techniques described by Aldrich. Everything can be clamped down and the transfer can be moved by N2 pressure. A 1.5 inch needle meant that she had to monkey with the bottle by tilting it into an awkward position. With the extra fuel in the hood that Morris described above, the transfer was rather precarious with a huge downside if she slipped up. Like walking the wire without a net.

  • Ken

    My opinion and the way I operate in my lab – – A syringe never should have been used in the first place for a reaction at the scale employed by Sangji. In my lab the tBuLi would have been transferred by cannula. The amount actually transferred would be determined by weighing the tared bottle (or reaction vessel) after the transfer. The amounts of other reagents required would then have been calculated based on the amount of tBuLi actually transferred. The reaction would have been scaled to nominally require the complete contents of 1 or 2 100 mL bottles of fresh material (typical bottle size available from the vendors), thus avoiding the storage of partially used bottles with punctured septa. I would never instruct someone to measure out such a precise amount (159.g mL) of something this hazardous when inherently safer methods could be employed instead.

    • gaussling

      Hi Ken,

      I have to agree with you. If you need to dispense up to,say, 20 mL max and have a Luer lock syringem then I’m fine with that. But as you say, the scale was well in excess of what should be syringed.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 112 other followers

%d bloggers like this: