A researcher was working with hydrogen storage materials in a laboratory. Several other researchers were working in adjacent laboratories.

The researcher had prepared a sample of aluminum deuteride, AlD3, by reacting lithium aluminum deuteride and aluminum chloride in diethyl ether. The actual composition/phase of the material synthesized was unknown, but the researcher had attempted to produce the gamma phase of aluminum deuteride. The synthesis steps used to produce the material were complete and the researcher attempted to seal the material in a glass ampoule for offsite shipment and analysis. The sample size was approximately 1 gram.

The ampoule with the sample had previously been placed under vacuum and had been isolated from the atmosphere. The process for sealing the ampoule was to first place the sample-containing ampoule in a liquid nitrogen bath to cool the sample down to near 77K. The ampoule was then removed from the liquid nitrogen and a torch was used to melt the neck of the ampoule to seal it. The ampoule was slowly rotated while heat was applied to the neck of the ampoule. Typically, this results in collapse of the ampoule, sealing the sample. However, in this instance, a bubble formed where the heat was applied and a hole formed in the ampoule. This allowed air to enter the previously evacuated container.

After about 30 seconds the ampoule “exploded” and glass from the container was sprayed outward. Some of the shards embedded into the researcher’s arms, face and torso (the researcher was wearing safety glasses, but no other safety equipment). Other researchers working in an adjacent laboratory heard the noise and came to see what had happened. The researcher was taken to the hospital and treated. Larger shards of glass were removed and the injured area was cleaned in an effort to remove the smaller pieces of glass. The researcher stayed home the next day, but returned the second day after the incident. There appears to be no permanent damage to the researcher and he/she is continuing work in the laboratory.

Incident Date
Apr 19, 2005
  • Laboratory Equipment
  • Glassware
Probable Cause
When Incident Discovered
Lessons Learned

Metal hydride materials of a composition which is not well characterized should be handled with procedures that assume a "worst case" for that class of materials, intermediates or precursors.
Laboratory procedures should be in written form and should be adopted only after performing a safety vulnerability analysis and adopting appropriate risk mitigation steps.
Working with small amounts of material does not provide assurance of safety.
The method described to seal samples that are highly reactive upon exposure to air is not recommended. An alternative method, that is now being used for packaging aluminum hydride samples for offsite shipment is as follows: Aluminum hydride powders (0.5 g - 1.0 g) are sealed in glass bottles with a cap. The bottles are sealed in a thick plastic bag under Ar using a plastic bag sealer in the glove box. The sample bags are then removed from the glove box and sealed under vacuum in a much larger plastic bag using a vacuum sealer. The purpose of the large evacuated bag is to contain any evolved hydrogen gas if the material begins to decompose and the internal bottle & bag rupture. It is a good idea to calculate the maximum volume of H2 (at 1-atm) that could be released by the sample to determine if the outer bag is sufficiently large to contain the evolved gas. The package is then inserted into a cardboard box of sufficient volume (about 1-liter) to accommodate the expanded outer bag.

Additional discussion about working with reactive metal-hydride materials in the laboratory can be found in the Lessons Learned Corner on this website and in the Hydrogen Safety Best Practices Manual.