More on the Sangji case, with key factors in the incident

The big news of the chemical safety beat this week was that the Los Angeles County District Attorney’s office filed felony charges against the University of California and UCLA chemistry professor Patrick Harran in the death of researcher Sheharbano (Sheri) Sangji. I’ve written a news story and two blog posts on the topic already; here are some additional links from yesterday and today (if I’ve missed anything, post it in the comments):

There’s also some Twitter discussion about whether there’s too much emphasis on the fact that Sangji wasn’t wearing a lab coat. First, I’ll note that there were several problems with Sangji’s experiment that day:

  • she was syringing the tert-butyl lithium (tBuLi) rather than cannulating
  • she was using a large, plastic syringe that would have been impossible to dry and difficult to handle
  • she was using a too-short needle and likely had to tip up the bottle to get to the liquid, making the syringe even more difficult to handle
  • she had an open flask of hexane in the hood

And when all of that led, somehow, to the syringe plunger coming out of the barrel and the tBuLi igniting everything nearby, Sangji further wasn’t wearing appropriate personal protective equipment (PPE) – a flame-resistant lab coat. Such a lab coat might not have prevented the fire entirely, but it would have slowed it and likely reduced how badly Sangji was burned in the time it took to put out the flames.

Another critical component to the incident is whether Sangji had prepared herself for the worst that could happen and at least mentally rehearsed the appropriate response: get to the shower. From the Cal/OSHA interview with the person who was in the lab with Sangji, it sounds like she panicked and ran around, fanning the flames.

All of this, of course, goes to training: Was she trained to know the best experimental procedure? Was she trained to consider and wear a flame-resistant lab coat? Was she trained to think through the experiment, what might go wrong, and how she should respond? Were her labmates trained in how to handle a panicking colleague with a shirt on fire?

The key point about lab safety is that it’s never just about one thing: It’s not just about minimizing use of hazardous reagents, or engineering and using the safest procedure, or using appropriate PPE, or knowing what to do when things go wrong. It’s about all of those together, with one component as the failsafe for another.

Edited to add Reddit Chemistry link.

Author: Jyllian Kemsley

Share This Post On


  1. Just to be precise, the problem was not that she was using a syringe, but that she was using a syringe to transfer such a large volume, right? Aren’t syringes the right tool for transferring, say, 2 mL (assuming the syringe is at least 5 mL and has been triple-rinsed with Ar, etc)?

  2. There is definitely excessive discussion about her lack of lab coat, but it isn’t focused in the right area. Odds are that there probably was a lab coat available for her to wear and she chose not to. However, odds are also good that it was probably a polyester/cotton blend, and therefore would not have offered any protection.

    The reason is simple; cost. You can get a cotton/poly blend coat for $20 or less. 100% cotton coats are at least double that. It doesn’t sound like a lot of money, but considering at the time of the accident it was extremely common for people not to wear coats at all, why spend the extra money on something people aren’t going to use? It’s easy to look back in hindsight and point fingers, but I can assure you that very few people were aware or thinking of this danger prior to the accident.

  3. There is one problem with her technique that has perhaps not been adequately discussed. With any size of a syringe, there is always a risk that the plunger pops out. This will happen if a) there is excessive pressure in the syringe or b) if the plunger is pulled with excessive force, or past the point of no return.

    In my view, all-glass syringes are much more prone to this than either plastic or so-called “gastight” syringes. Some people have claimed that all-glass syringes would be better because you can dry them properly. This is true, but then again glass is a material that adsorbs water on its surface, whereas polyethylene (the material used for plastic syringes) does not, hence there is virtually no need for excessive drying! As such I would recommend using either gastight or plastic syringes for transfer, but definitely not in 50 mL scale.

    As such I believe the technique (just the technique, I am not commenting on the lack of PPE) Ms. Sangji was using was basically OK for a smaller scale experiment, but what was not OK was to use a 60 mL syringe for the transfer. These plastic syringes are so tight to pull that they require excessive force to be filled – which, if I may speculate, may have directly led to the accident.

    This goes to show that the danger lies in the details – and scaling up is more dangerous than most people realize!

  4. I wonder what type of plastic syringe was used? I use disposable plastic syringes (without the black rubber at the bottom of the plunger – these are no good for chemicals) and I find that they all have a stop ring on top which makes it hard to pull the plunger out.

  5. I just looked at the picture, it appears to be a good airtite disposable luer lock syringe.

  6. I am very sorry for this tragic incident, but as career scientist with over a decade of experience in working with hazardous substances, I can tell that she got overconfident. You do not work with substances that can ignite from air contact in open hood, with syringe that has a removable plunge and without PPE. You may do the same procedure 1000 times with no issues and get the idea that long and sometimes redundant protection routine is something you can forget about, but there is always that 0.000001% chance that it will go wrong. And that is exactly what happened, imho.

    My heart goes to poor girl family, but blaming professor here is wrong.

  7. @DoubleEM – All the available evidence suggests that Sangji was hurt the second time she handled tBuLi. Where do you get the idea that she did it many times with no problems?

    And with respect to this: “You do not work with substances that can ignite from air contact in open hood, with syringe that has a removable plunge and without PPE.” Who do you see as responsible for (not) ensuring that she had the training to do the correct procedure?