Skip to main content

Beryl Lieff Benderly

California Investigation Report Explains What Went Wrong for Sangji

Yesterday we published an item about the 2009 report on the investigation by California’s Division of Occupational Safety and Health into the death of Sheri Sangji. I suspect that some readers were frustrated that we didn’t describe what went wrong during Sangji’s fatal attempt to transfer a quantity of tert-Butyl lithium. The report’s author, Senior Special Investigator Brian Baudendistel, provides an account that goes on for pages. The basic issues come down to missing training and inappropriate equipment. Of course, these issues are related because a properly trained worker would know to use the right gear. 

Here’s a summary of what the investigation found.

Because tert-Butyl lithium material is extremely hazardous and ignites on contact with air, the report notes, each bottle comes with extensive warnings and the manufacturer’s specific instructions for safe handling. “There are two basic methods of
transfer…(1) the syringe method and (2) the cannula method,” Baudendistel explains. “Each procedure requires the use of laboratory glassware that is subjected to oven drying to remove residual moisture prior to use. In addition, both the reagent and reaction flask must be purged and slightly pressurized with dry nitrogen to displace any air within the system…. Further, the reagent bottle must be clamped in the fume hood to secure it from displacement during reagent transfer.”

Next he quotes the manufacturer’s technical instructions: “Small quantities (up to 50 ml)…may be transferred with a syringe equipped with a 1-2 ft long needle.” The long needle permits the procedure to take place
without the worker having to tip either the bottle or the storage flask. “In general,” the instructions add, “a syringe should only be used for a single transfer.” Reuse “can result in plugged needles and frozen syringes….”

For transfers above 50 ml, using a
cannula–a tube “with a syringe needle at each end”–“is preferred” to a syringe with a plunger because the cannula provides “essentially a direct connection between the respective vessels and minimizes the exposure to lab personnel,” Baudendistel writes.

Neither method is perfectly reliable; either “can expose personnel to the risk of fire and/or explosion,” making “appropriate planning and protective measures” — including protective appare l– “a prerequisite irrespective of the method chosen.”

As the
transfer gets underway, “the nitrogen pressure is used to slowly fill the syringe with the desired volume,” the manufacturer’s technical advice continues. “Note that the nitrogen pressure pushes the plunger back as the reagent enters the syringe.  The plunger should not be pulled back as this tends to cause leaks and create gas bubbles.”

At approximately 1 P.M. on December 28, 2008, working “unsupervised” in a fume hood “crowded with apparatus” that made the work area “somewhat restricted” and clad in a synthetic sweatshirt and no lab coat, Sangji began her attempt to “transfer 159.5 ml of t-Butyllithium, in transfers of approximately 53 ml each, using a 60 ml…polypropelene syringe.” Being plastic, the syringe could not be oven-baked. It also was too small to accommodate safely the amount of material being transferred. According to the “prevailing scientific literature,” Baudendistel explains, the syringe should be “at least twice the size of the intended transfer.”  In addition, “it appears that Victim Sangji attempted to make multiple transfers of t-Butyllithium using the same syringe”–“a
common practice in Dr. Harran’s laboratory.” And probably most important of all, the syringe was “equipped with a…2 inch long needle, as opposed to the 1-2 foot long needle specified by” the manufacturer. This “prevented Victim Sangji from properly clamping
the…bottle.” Because the needle was “too short to reach the necessary amount of reagent if inserted vertically into the top of the reagent bottle,…Victim Sangji was required to either hold and tilt the bottle with one hand and pull the on the syringe plunger with the remaining hand to make the withdrawal, or lay the reagent bottle on its side on
the bench top…and attempt to complete the transfer with one hand while stabilizing the reagent bottle with the other.” Although “contrary to the express warnings issued by” the manufacturer, Baudendistel writes, “manual manipulation of the plunger [was] confirmed as an accepted practice by Dr. Harran.”

As Sangji attempted the transfer, “the plunger of the syringe became dislodged from the syringe barrel, causing the reagent to be released.” It instantly burst into flame, igniting her highly flammable shirt.

Why did Sangji go about an already dangerous task in a way that needlessly
multiplied the danger? Because, it appears, no one had ever taught her otherwise, apparently. “Dr. Harran admitted that the [manufacturer’s instructions were] used as a ‘general reference’ only and that training relative to the handling of t-Butyllithium was based on ‘knowledge’ passed down from one generation of researcher to another. Dr. Harran also admitted that he never discussed with Victim Sangji the risks associated with” the task she was undertaking.

“Dr. Harran claimed that Victim Sangji had been properly trained to handle t-Butyllithium by one of his…postdoctoral researchers, Dr. Paul Hurley,” Baudendistel continues. Hurley, however, “did not have any specific recollection of the actual guidance offered to Victim Sangji, if any. Aditionally, Dr. Hurley did not have any specific recollection of providing ‘formal training’ to Victim Sangji relative to the syringe transfer method….” Harran, furthermore, “admitted that he…had never attempted to determine if Dr. Hurley had actually provided any guidance or instruction to the Victim, relative to the transfer and handling” of the material that would take her life.

Hurley “confirmed that he did not follow” the manufacturers safety instructions “and did not believe that he had ever read” them. But he did watch Sangji transfer a much smaller quantity of t-Butyllithium some 2 months before the fire. On the fatal day, the report notes, “Victim Sangji employed many of the same improper techniques used by Dr. Paul Hurley, which suggests that Dr. Hurley had provided some level of guidance during the Victim’s completion” of that task in October.

“While UCLA delegates much of its responsibility for worker safety to Principal Investigators,” the report observes, “the University did not require PI’s to attend safety training prior to conducting research in
their assigned lab, nor did the University make any effort to evaluate PI’s fitness or competency to comply with and enforce applicable workplace and laboratory safety regulation prior to supervising employees.”