Update on death at Tsinghua University

An explosion in Tsinghua’s Ho Tim chemistry building killed researcher Xiangjian Meng on Dec. 18. Credit: Rolex Dela Peña/EPA/Newscom

An explosion in Tsinghua’s Ho Tim chemistry building killed researcher Xiangjian Meng on Dec. 18.
Credit: Rolex Dela Peña/EPA/Newscom

From my story in Monday’s issue of C&EN:

“We are deeply saddened by the accident and loss of a good postdoc scientist,” Tsinghua chemistry department chair Xun Wang told C&EN.

“According to the investigation of the government police department as well as our own investigation, the tragedy was caused by the accidental explosion of a hydrogen gas cylinder,” Wang said. As of Dec. 30, the blast’s cause was unclear.

According to a post by the Beijing Administration of Work Safety, a Beijing vice or deputy mayor led some sort of safety inspection at Beijing University of Chemical Technology on Dec. 30. From the google translation of the post:

Wang Ning, deputy mayor stressed that one should draw profound Tsinghua University “12.18” explosion lessons, school leadership to further enhance the understanding of the management of dangerous chemicals laboratory. Colleges and universities are the place to nurture talent, but also to pay attention to people’s safety, especially teaching laboratories and research laboratories using hazardous chemicals, storage security is a top priority. Second, the schools to organize the pre-holiday special inspection responsibility to the people, the use of hazardous chemical storage to conduct a comprehensive clean-up, focus on examination of system implementation, test personnel training, check targeted contingency plans and emergency response, security. Third, the City Board of Education in accordance with the requirements of the Ministry of Education on College lab safety management, strengthen safety supervision system in Universities, in close coordination with safety supervision, public security departments, from management mechanism, team building, development of the system, rating, personnel training and risk analysis, study and formulate targeted measures; summing up and promoting some good experience and practice, work together to promote the city’s university laboratory safety management of hazardous chemicals, and strive accident does not occur.

So 1) colleges and universities need to pay attention to people’s safety, especially in teaching and research labs; 2) there will be lab clean-ups and inspections, and 3) education authorities are going to pay closer attention to safety. My question: Will any of the government or university responses lead to a culture that fosters working safely as opposed to a culture of compliance?

h/t to C&EN’s Jean-François Tremblay for the Beijing Administration of Work Safety link.

Author: Jyllian Kemsley

Share This Post On


  1. “…accidental explosion of a hydrogen gas cylinder.”

    Nope – I’m not buying it. This keeps getting stranger and stranger. I don’t think we’ll ever know the real cause.

  2. Harry, why do you think it’s not a hydrogen gas cylinder?

  3. I can’t envision a situation where a hydrogen gas cylinder would explode “accidentally”. Hydrogen gas, yes, but not the cylinder. The most likely cause would be misuse or misapplication of the hydrogen, with the ignition source being either static electricity or open flame. With an LEL of 4%, it takes a fairly significant concentration of H2 for ignition and a lack of adequate ventilation.

  4. I am guessing it may be due to pressure vessel rupture (hydrogenation) due to uncontrolled pressure as they mentioned “the schools to organize the pre-holiday special inspection responsibility.” It means the reaction was left running during the weekend and something went wrong (uncontrolled heating and pressure). Hydrogenation catalyst (Pd-C, Raney Nickel) in addition to solvent and hydrogen can give a huge fire PLSUS flammable solvents and reagents near to reaction set-up; eventually high temperature surrounding the tank can rupture the main H2 tank.

  5. Or the “pre-holiday special inspection” means that authorities want labs to be cleaned up and expected inspected before the Chinese New Year (Feb. 8) and associated Spring Festival.

  6. Flashback arrestor is easily overlooked if people are not trained and/or not equipped properly. I believe there are plenty of instances of explosions caused by failure of flashback prevention, and failure to include flashback prevention. Are those “accidents?” Doesn’t matter; people call everything unfortunate a FREAK ACCIDENT no matter how predictable it was and preventable it would have been if some training and good oversight had been involved.

  7. Well, I don’t think anyone has called this a “freak” accident. Most likely it was an incident that could have been anticipated and prevented given appropriate training, equipment, and oversight. At the same time, it was probably also accidental in that no one intended to set off the explosion and fire.

  8. The rupture of H2 cylinder from valve or cylinder collar will land the RUPTURED cylinder to a different destination (rocket effect), or will break upper floor with huge damage. May be the building was very strong, or it ruptured from its body.

  9. Jyllian:

    Sorry for the slow response – I had email notification turned off (corrected).

    Russ pegged it. Cylinders don’t “explode” without some underlying cause. The cylinder may over-pressurize causing a rupture but there must be something that contributes to the over-pressurization like a fire. Or you “can” break off the cylinder valve to release the contents into a room, but then again, there is an underlying event to that.

    Emptying the contents of a 2000 psig cylinder into a room may cause a deflagration after it finds an ignition source. Additionally, releasing high pressure hydrogen into the room through an orifice (such as an OP plug) may also cause a fire, as hydrogen has a reverse Joule-Thompson effect (i.e. it heats up on expansion through an orifice instead of cooling down as most compressed gasses). The effect is prominent enough to cause ignition.