Thursday, June 4, 2015

UC Berkeley safety report: student injured during explosion at UC Berkeley

No words needed.
Credit: UC-Berkeley EH&S.
Via a tweet by longtime chemblogosphere denizen Curt Fischer, a EH&S report from UC-Berkeley: 

Lesson Learned - Chemical Explosion Causes Eye Injury
What happened? 
A graduate student researcher was working at a laboratory bench synthesizing approximately one gram of diazonium perchlorate crystals. The student was transferring synthesized perchlorate using a metal spatula when the material exploded, sending porcelain fragments into his face. The fragments shattered the lenses of his eyeglasses and lacerated his left cornea. 
A researcher in an adjacent room assisted the student to the eyewash and called campus police. The student was taken to the hospital where he underwent surgery on his eye, and treatment for several facial lacerations. He was released from the hospital that same evening. 
What went right? 
The student was wearing a flame-resistant (FR) coat and nitrile gloves, as called for in the relevant SOP. 
The student had previously read and signed the relevant SOP, and had completed EHS 101: Laboratory Safety Fundamentals training. 
Emergency protocol was followed once the incident took place (eye wash, contacting emergency personnel). 
What should have been done differently? 
The student was not wearing ANSI-approved safety glasses as is called for in the SOP and in UC policy. 
The student was alone in the room while working with highly hazardous materials. 
The work was being performed on an open benchtop, without the use of a fume hood sash or blast shield. 
The transfer was performed using a metal spatula, while the SOP calls for use of a non-metal one.
Click here to read the entire report; it's worth your time. I have to say, I'm pretty shocked that an incident like this could have happened for a number of reasons:
  • Post-Sangji, I was under the impression that the UC system came down pretty hard on PPE issues. I guess not hard enough. 
  • Perhaps I'm naive, but I think the Preston Brown case is plenty instructive to energetic materials work and super-duper relevant to this case. Why weren't those lessons learned (don't make too much, wear the right PPE, use the right tools) in this case? 
I'd like to know when this case happened - I see that the report was issued "May 2015." That suggests that it's relatively recent, but who knows? 

Finally, as I am wont to do, I have to ask - where was the PI in this mess? How is it that the graduate student managed to do this? 

Academic chemical safety - we're not there yet. 


  1. 1 g of a diazonium perchlorate seems like an awful lot; maybe they were doing phys org studies of their reactions, but other than that, what do you use that much of a diazonium perchlorate for? (If I were using it as a reaction intermediate, I'd be happier to use another counterion if I could, or generate it in situ instead if I couldn't, and if I can't do either, whoever's going to follow your method (or you) had better be desperate.)

    Wearing regular glasses around 1g of something likely to be explosive also sounds like a really bad idea. I never liked goggles (school paid for lab-safety goggles, though they wouldn't have worked against splashes to the face) but if I don't have good glasses than goggles would seem like a really good idea.

    This just sounds like a bad idea - not as bad as 10 g of a nickel perchlorate, but bad enough.

  2. It really seems like the PI should be aware that his students are not wearing approved safety glasses/goggles. Part of the responsibility of managing a research group includes creating a proper safety culture. Wearing proper safety glasses/goggles is very basic and probably the most obvious rule that should never be broken.

    1. statistically "his" is the correct assumption.

  3. The PI was probably relaxing in their office, assuming that since the Grad Student had signed an SOP stating that he would wear PPE and use a plastic spatula, that the he would actually do those things, as well as following common-sense guidelines like not working alone and keeping explosive compounds behind a blast shield.

    The important thing to remember is that this is, entirely and criminally, the PI's fault, and that the individual who broke all the rules bears no responsibility because he got hurt.

    1. I'd be happier with the implication of the sarcasm if PIs were ever actually held responsible for what happens in their labs. Don't really have to worry about that, though. It's not like training people is their job or anything...

      (I don't remember people checking, but since school actually paid for safety glasses, I'm pretty sure they intended them to be used. I would imagine that part of the point of all that overhead money a school gets is to manage overall safety requirements, of which safety glasses would be one.)

      I assume the point of filling out the SOP was to make you think hard about what you're doing. (Part of the point of safety infrastructure is to make it hard for you to avoid knowing about what you're doing, and to make it harder to hurt others while doing it.) If you didn't think about what you were doing after that, it's unlikely that someone else is going to make you think about your actions. There are other problems as well, but if you don't know enough to know that you might want to be careful after filling out those forms (and what being careful in your case means) and don't know or care enough to ask for help, then you have a problem. And most of the blame for whatever happens next.

    2. "The important thing to remember is that this is, entirely and criminally, the PI's fault, and that the individual who broke all the rules bears no responsibility because he got hurt."

      I don't think I have enough evidence to make a judgment either way, or to begin to slice the blame pie.

    3. Oh, I agree. Once a research group gets to the size that the PI doesn't know what each researcher is doing every day ( ≤ 8 people, maybe?), they're just asking for it. Is it not the best way to train students? That depends on the student. Is it criminal? I don't think so, but I only work with undergrads.

      Re: PI responsibility, they're not accountable for the psychological abuse they subject their underlings to, so why should they be for safety? It's like jailing Al Capone for tax evasion!

      I agree about the forethought, but the other point of signing the SOP is to document that, in a situation like this 1) there was an established, safe procedure; 2) the individual was aware of this proper procedure; and 3) the individual willfully chose to disregard the procedure. We didn't have that in the Sangji case.

    4. I guess I assumed that the SOPs were unusual - if they weren't that would dilute the effect of them on how people in the lab perceive and account for the risks of a reaction.

      On the other hand...a gram of a diazonium perchlorate seems like something I'd want to know the risks of beforehand.

  4. As a grad student at UC Berkeley, I can say the following:
    1. CJ, in response to being under the impression that UC came down hard on PPE issues after that incident, all I can say is the response was a huge joke. We now take this EHS 101 training, which is just an agglomeration of all of the trainings that existed prior to the incident, performed online one time with quizzes that you cannot fail (literally, if you get a wrong answer, it just says try again until you get it right. Everyone gets 100%). No one enforces lab coat or safety glasses requirements or really any other safety standards. In my entire time there and at a previous institution as an undergrad, I've never been in a lab that was inspected by OSHA.

    2. The SOPs are a joke. The requirement that everything have an SOP came into effect after the incident and EVERYTHING had an SOP, including things like sodium chloride, sodium hydroxide and the like. In many cases, we had to write these ourselves which was onerous and no one did thoroughly. This resulted in everyone in the lab signing a document saying that they had read every SOP when that was clearly not true.

    3. The PIs are nowhere in this. Almost none of them could pass a quiz on correct safety procedures in their own labs, much less enforce the procedures.

    4. Asking EH&S officials whether something is safe tends to result in a long winded answer (NEVER in writing via e-mail, only in person) that doesn't end with a "yes" or a "no". They are uninterested in actual safety, only the universities liability for incidents. Thus the avoidance of definitive answers and paper trails about safety questions, but keen attention to whether someone has signed the documents saying they have read the SOPs.

    We're not only not there yet on academic chemical safety, we're nowhere close.

    1. Still probably a heck of a lot better than the Chemistry labs in China and India. No wonder the PHD's there want to move to the US.

    2. When it comes to safety, US academia is a hell of a lot closer to China and India than it is to US industry.

    3. >4. Asking EH&S officials whether something is safe tends to result in a long winded answer (NEVER in writing via e-mail, only in person) that doesn't end with a "yes" or a "no". They are uninterested in actual safety, only the universities liability for incidents. Thus the avoidance of definitive answers and paper trails about safety questions, but keen attention to whether someone has signed the documents saying they have read the SOPs.

      I"m coming at this from an EHS point of view, but "whether something is safe" is not an answerable question. My experience is that when chemists questions are usually asked in passing without the context necessary to answer them - "yes" and "no" usually require a lot more details than the chemist has at hand.

      I think it's also important to remember that the University's liability is how academic science is sheltered from the kind of public concern that arises when they hear about home labs like the one at UC Davis or meth labs. This shelter is eroding rapidly, particularly in biolabs, under the pressure of emerging technologies.

      If it was a simple yes or no question, you wouldn't need to ask an EHS person.

    4. If you ask EH&S about safety they will pull up the MSDS and SOP and... As I was always taught go find someone that knows what they are doing. Especially at Berkeley, go talk to a senior graduate student or postdoc with significant experience. Ask a number of people if they have advice on how to do something safely.

      I will second that unfortunately I see the SOPs as a bit of a joke. They are just today's version of an MSDS. Everything is toxic and dangerous, don't eat it. When everything is listed as being so horrific it desensitizes people who then go on to think "Oh yeah KOH had an SOP, I used it and it was nothing... I also signed that SOP for tBuLi and KCN..."

    5. EHS people: your job when you cannot science.

    6. @ KT

      When it comes to safety, US academia is a hell of a lot closer to China and India than it is to US industry.

      I recently graduated from a big ten university and did a post doc at another large university. I am currently working in pharma, and can say that my experience with Universities in the US is different than yours. Safety was a major concern and taken seriously. But I will admit a lot of the enforcement was brought on by the senior students and post-docs.

    7. Maybe it has something to do with half the students.

      When I did my engineering courses I frequently felt like I had somehow mistakenly gone on study abroad to Mumbai.

  5. And frankly, we won't get there until the proper incentives are in place. Until the universities and professors are held accountable financially, socially, or otherwise, the culture of academic "safety" will continue to flourish. If the same incident happened at Big-Time Chemical Company, how do you think the situation would have unfolded?

  6. On the topic of PPE and safety glasses, is there a consensus on the use of contacts in the lab? I know there are prescription safety glasses with side shields (I use a pair) but they seem to offer less protection than normal safety glasses, and far less than safety goggles of course. It would be nice to wear contacts with normal safety glasses/goggles but there are various horror stories about contacts and solvents, and removing them in case of emergency would be almost impossible.

    1. Somebody reviewed the literature on wearing contacts in the lab in the mid-2000's, and discovered that the old warning about not wearing contacts in the lab was an old wives' tale that kept getting repeated, with no basis in facts. I was really surprised when my super-cautious big chemical company employer started allowing contacts in the lab about ten years ago, after not wearing contacts was one of the few safety rules we had in grad school.

    2. NIOSH has guidelines for wearing contact lenses:

    3. Excellent, thanks to both of you for the information!

  7. It bothers me to see blame being thrown around indiscriminately so let me add a few thoughts. As a student at a UC school, I have seen several PIs, including my own, who were proactive about safety before the Sangji case and were genuinely concerned about students' safety. We were inspected by Cal OSHA as part of the settlement and I can attest to the huge amount of useless paperwork that was signed for legal purposes, but there was also a genuine push (at least at my school) to improve the safety culture. Please, use sarcasm and blame where it is earned, but some PIs at big schools do actually care about students as people and take their mentorship-role seriously. They are not, however, omniscient or omnipresent (although some of the worse PIs like to pretend they are.) Despite my PI actively calling-out people who do not use proper PPE, when the boss isn't around, some labmates still wear whatever is comfortable for them when they are not running a reaction that worries them. To paraphrase one of my friends, we, as grad students, have become pretty cavalier about chemical safety and don't like to follow rules when they slow down our research progress or inconvenience us. At the end of the day, if the people running the reactions don't take safety seriously, accidents will continue to happen.

    1. As long as grad students are judged solely, and harshly, on research productivity, I see no improvement in this area. I used to advise new hires at Big Chemical Company that a safety incident will get you in more trouble than poor productivity will. I don't think a punitive culture like the one at DuPont is a good extreme either, but the way it is now, a grad student has every incentive to cut corners on safety if it will help him/her get more experiments done.

  8. "The chemical activity of perchloric acid and the fact that, under certain conditions, it is explosive, are matters of common knowledge, it is also generally known that many of the diazonium (diazo) salts such as benzenediazonium nitrate, C4H4N:NNO2, are likewise explosive and their instability is often very great. It might be anticipated, therefore, that diazonium perchlorates, if they could exist at all, would prove to possess an unusual degree of energy and that their study, although it might be attended with danger, would yield results of considerable general interest."


    So begins a Science paper on the subject (Science 1 February 1907: Vol. 25 no. 631 pp. 193-194 DOI: 10.1126/science.25.631.193 ). This author definitely has the right attitude. General interest! Fun for the whole family!

  9. Main thing here is the widespread belief that eye glasses are the equivalent of safety glasses. Unfortunately, the best safety practices for graduate students are your own personal self-preservation instincts. Many times these instincts are discarded to accommodate the pressures of taking the quickest route to get things done. People are learning the wrong lessons in graduate school (quickness over quality) and situations like this are examples of it.

  10. I just posted this at The Safety Zone, I'm going to repeat it here:

    What a mess, where to begin?

    Diazonium salts are generally explosive (the report does not indicate which one was involved). Perchlorates, especially organic, are explosive. A diazonium perchlorate??!! Absolutely explosive. Did the PI know this? If so, what in the world was he/she thinking?!

    After reading this post I turned to Google in an attempt to learn more. The first hit was a cached Word document, apparently a initial draft of this incident report. That draft indicates that the affected individual was a first year graduate student. This is missing from the final report, so either the initial draft was wrong or it was purposely omitted from the official report. It would be good to know which is correct. Here is the cached draft:

    The report refers to an SOP, but a copy of the SOP is not provided. It can be critiqued, however. For example, the report states that the student was wearing a lab coat and nitrile gloves, as called for in the “Standard Operating Procedure for Potentially Explosive Compounds”. Really? A lab coat and nitrile gloves are adequate protection against shrapnel? Unless you have the same PPE used by bomb disposal squads, flesh and bone should be nowhere near explosive compounds. Everything must be done behind a blast shield and using the correct tools designed for explosives work. The student is lucky he was hit in the glasses, a few inches lower and he could have been hit in the carotid artery. PPE is your last line of defense. In the case of explosives few if any labs have PPE capable of providing that protection. Perhaps this is stressed in the SOP, and the report does indicate that the student failed to use a hood sash/shield, but this point is buried several bullets down in the report when it should be at the very top.

    Read through the document referenced in the report – I did – and see if you can find any guidance whatsoever on the handling of explosive compounds. I can’t find any. It’s the one titled “Campus Guidelines for Potentially Explosive Chemical Safe Storage and Handling” at the bottom with the pdf hyperlink. It’s woefully inadequate for dealing with the type of chemical involved in this incident, to the point of being irrelevant.

    The lesson learned here is NOT about eye protection. It’s much, much more than that.

    1. Excellent point - standard chemistry lab PPE is meant to protect you from chemical splashes. Unless you're working with energetic materials, it's unlikely you'll encounter an explosion outside of teaching demos, and ordinary PPE isn't designed for it.

    2. 1) If the student was a first-year, then they weren't being smart, but a lot more of the blame needs to be elsewhere than on him (where I put it, originally). Particularly with a gram of a likely explosive.

      2) There's methods to deal with explosives, but I imagine generating the compound in situ, not at all (different salt?) or on smaller scale would have worked better - oversight might have helped.

      3) In a game of CYA, the person with the least power loses.

    3. Another UCB student here. The student in question was a first year, but details are scarce even here. While this doesn't excuse them or the PI, it may explain why such a cavalier attitude was taken toward such an obviously explosive compound. I will say that my PI here has been very clear that proper PPE is extremely important and anyone who gets caught not adhering to proper safety procedures gets chewed out accordingly.

  11. This is Curt F. I beginning to think that the most important job a PI could do for new students in regards to safety is to tell them that EH&S, all the accompanying quizzes, SOPs, forms, etc. have very little to do with safety. If I were a new PI, I'd tell new students that they should be terrified every day they step into the lab. Every day they could kill or injure themselves or their colleagues. And since they are new and inexperienced, they are more likely to do so. I'd also say that when it comes to safety, you should assume the worst about everyone. The EH&S officer's primary concern might be keeping you safe, but it probably isn't: keeping their job and getting promoted may well be first on their agenda. My job as a PI is to get grants and write papers, the sooner the better. So don't assume that I will be keeping you safe. The only hope for safety starts with you.

    I also don't understand why everyone is already calling the PI a criminal in this thread. I agree with CJ that it's far from clear where culpability lies. First-year (?) graduate students are adults, and apparently this one signed his name to a procedure that said he would not use metal spatulas and would use proper safety equipment. Unless evidence emerges that the PI fostered a culture of ignoring or disregarding SOPs, it's hard to understand how the PI could be a criminal. That's especially true if the PIs have in general forgotten all their lab technique!

    That said, the other points that this was a very large amount of explosive compound to be working with, and that the SOP's recommended PPE were inadequate for it, are very good.

    1. Employees in industry are also adults.

  12. On the one hand, I think it's instructive to highlight safety incidents like this, and I think it's good that CJ and other bloggers continue to do so, but, on the other hand, the comments are always full of people jumping to conclusions about who is to blame.

    Presumably, as scientists, we should gather all relevant data and then form reasoned conclusions based on that data. Instead I see people jumping to condemn either the PI, or the student, or the EHS department, etc. The blame could arguably belong to any of the above, or to all of the above.

    I don't know if it's productive for the blogosphere to try to find blame in these situations, but here are some of the factors I'd be interested in:

    --Was this a lab that has a program focused on these diazonium perchlorate salts? If so, then I'd expect the PI to ensure that all group members were thoroughly aware of the hazards and appropriate handling practices. If not, and the student had just run across something in the literature and said, "hey, I can use this for my work," and then just went ahead and made it, I think it's a little more complicated. It is possible that a young grad student wouldn't know that this sounds dangerous. Most of us have a kind of learned intuition that would causes our brain to go "danger! danger!" when we hear something like "diazonium perchlorate." Presumably we all learned this somewhere, but where did we learn it? Things like this should be highlighted in the safety training everyone takes before they join a lab--what classes of chemicals should raise flags and require additional precautions--but I find a lot of safety training is too vague, making everything sound equally dangerous, causing people to become complacent and not recognize things that carry more significant risks. More experienced labmates are also often a source of knowledge. But no matter how much training you have though, I'm sure there will always be some gaps--no one knows everything. All researchers, no matter how experienced, should be responsible for understanding the limits of their own knowledge, and if they're doing something they're not familiar with, they should look into it and assess the risks before they start.

    --Working alone without proper safety glasses. Again, the PI and the institution should be establishing a culture where this isn't OK, and if they aren't doing that, then something needs to change. But maybe that is the culture and this person is the anomaly who went against it for whatever reason.

  13. And the circlejerk of "academic safety is lax" continues. Some student is an idiot and decides to handle an energetic between porcelain buchner and metal spatula (friction anyone?) and it goes off. And we play the blame the PI game and wish for the crushing inefficiency of the industrial SOP.

    This is why I left chemistry. Either gotta deal with shitty funding in academia, or take it up the butt with safety regs in industry.

    1. Which is it? Did you leave a job in academia because of shitty funding? Or did you you leave an industrial job because of safety regulations?

    2. both. disliked whoring myself in academia for funding, metaphorically speaking of course. Went to industry. Had to get approval from others to do anything at the bench. Want to switch recrystallization solvents? that is a new SOP. Was spending equivalent time begging for permission to do science, as actually doing science. So I said fuck it and went to work at a hedge fund.

    3. Decided to get out of the whoring business?

    4. Metaphorically speaking, of course.

    5. if I'm going to be whoring, I'd rather be a well paid whore.

    6. I thought that's what you'd say. You know, I think you may have landed on the wrong site... Google can probably tell you where to find Financejobber.

  14. Signed SOPs are a very poor substitute for proper one-on-one training in the academic research lab. More useful documentation in an academic setting would be documentation that "On (date/time) (name of senior researcher or PI) offered instruction on (hazard) and verifies that (student/ junior researcher/ new lab member) has demonstrated competence to carry out the named procedure safely without direct supervision. (signatures of both researchers)" Each lab would then have a list of it's hazardous reagents/procedures/equipment and each lab member would have a posted training log of which hazards they are have been trained to handle safely.

    1. I prefer a single rule stating
      "don't know what you are doing? ask someone who does!"

  15. I am a bit late to the party - sorry - and I am confused. Does UC-Berkeley SOP permit work with primary explosives:
    - after completing just EHS 101
    - in non-conducting PPE and on non-conducting surfaces
    - without written experimental procedure that was reviewed and signed off by another competent chemist
    - without apparent control of the quantities of synthesized material (near 1 g smaller blast shields may become projectiles themselves)?
    If the answer is YES to any of the above they will continue to lose students. If the EHS SOP does not specifically describe permitted conditions for fork with primary explosives this kind of work is simply not allowed at any quantity.

    With the primary explosives the question is not "whether" they will go off. The question is "when". The setup and training must allow for the survival of the chemist with an intact number of digits, eyes, etc.

  16. Auntie MarkovnikovJune 9, 2015 at 7:27 AM

    (I wrote the above post AnonymousJune 5, 2015 at 8:07 AM)

    There isn't a lot to go on from what was published about this incident. I'll venture an opinion, though: This incident is an example of incompetence on the part of the UCB EH&S, the UCB Chemistry department, and/or the PI involved. It is clear to me that UCB EH&S assumed that their document titled "Campus Guidelines for Potentially Explosive Chemical Safe Storage and Handling" adequately prepared students for this work. Take a look at that document and see what you think - the link is provided in the report. It's worthless. I'll wager the unpublished SOP is also inadequate for work with this type of material, and therefore the responsible entities are ignorant of the risks and hazards involved, and thus incompetent to oversee or train someone to do the work. They threw a worthless training document and a likely worthless SOP at the student, assuming that enough paperwork would cover the bases. The incident happened and they zero in on incorrect PPE as the culprit, failing to understand the root causes of why the incident happened. Again, my opinion, but this is an example of the Kruger-Dunning effect (

    Apologies for the tone of my posts, but the snarky writeup from the UCB EH&S department really pisses me off, especially since I think they bear a lot of the responsibility for this incident. (FYI I am a long time industrial R&D chemist and I constantly battle management and our safety department over their over-reliance on paperwork and SOPs)

  17. Thank you Auntie for the D-K ref. I can finally put a name to what I have been experiencing most of my life :).

    I followed the links from the report and tried to put myself in the roles of the student, the PI, and the EHS person. I tried to divine who contributed what to the accident investigation and report. I also tried to use the UCB EH&S system as advised in the report’s conclusion. Finally, I tried to reconcile this with the idea of actually graduating from a Ph.D. program in relatively intact form – at least physically intact.

    I can report two conclusions:
    - Everyone (students, PI, EHS people) is doing their primary job reasonably well.
    - The system is failing and unworkable.

    To me the main failures of the system in these five areas:
    - Promotion of adversarial relationships in the triangle EHS-PI-student (“his fault, not mine!” and “I will write you up!”).
    - The insistence on finding a single, strong “root cause” of any adversity.
    - The increasing complexity of all activities around research. Complex environment = complex and difficult to forecast hazards.
    - Use of tools that falsely appear to address discovered problems. They may be “the best we have” but they were not designed for their current use and cause more problems than fix.

    I can elaborate on these points in a separate reply. Let me know.

    I am somewhat optimistic that we can make progress in lab safety. None of that is going to happen quickly, though, and someone’s life and limb will suffer in the process.

    1. Oops! I did have a fifth cause, but I decided that it was redundant.....

  18. What about expelling any student found in the lab without wearing proper PPE?

    1. This would seem like a natural reaction. The EH&S mandates proper PPE for everybody's benefit and when a student declines that benefit and exposes the school to financial losses he should go.

      The student should probably be disciplined. I am pretty sure that expulsion on a first EH&S offense is not the right punishment.

      There are two parts to reactive discipline:
      - determination that an offense was committed;
      - determination of proper response.

      The accident had already happened and post-accident review found that the student did not have the mandated PPE. The review does not address whether the mandated PPE would provide good protection (perhaps not) and does little to review the selection of engineering and procedural controls. The PPE needs to protect the student only when something unexpected happens and selected engineering and procedural controls fail. When the experiment runs as expected the PPE does not need to protect the student because there is no exposure. The risk-based selection of procedural, engineering, and PPE controls (in this order) is mandated by OSHA.

      When I consider the procedure my conclusion is that everything that happened was expected. The procedure used was almost certain to result in an explosion. Risk analysis was apparently not done prior to the experiment, so engineering controls were never selected, never mind the proper controls. The not used default control (exhaust hood) probably would not have protected the student anyway. These are failures of the PI first and the EH&S as a very close second. The student was culpable, but to a minor extent.

      In the last part I considered the idea that a university is just like a corporation and should act like one, i.e. expel a student who contributes to a loss. The primary, chartered role of a university is to educate. Making money off grants must be a distant second. When you expel a student for a breach of an internal regulation you break that education cycle for the student and for whoever is watching that process. This anti-educational effect is even stronger when the regulation in question was so poorly written and executed.