Can capacitor burst be dangerous?

If a capacitor explodes due to a revers voltage being applied, then toxic gases and corrosive chemicals can be expelled and if they are ingested or land on skin or clothing can produce burns not only from the chemicals themselves but also because they were superheated in the explosion itself.  You will need to check the manufacturer's data sheet and material safety data sheet (MSDN) to know exactly what those chemicals are.

Depending on the size of the capacitor it may or may not have a ventilation mechanism.

When bringing up newly designed boards, there is always the possibility of a capacitor having a reverse bias applied.  This can be due to an improperly defined footprint during the design phase or because it was installed incorrectly during the manufacturing/assembly phase.  Either way, it is always a good idea to verify by physical inspection that all such capacitors are installed correctly and trace the PCB connections in the gerber files to validate the expected bias is applied.

Even with taking these precautions, some capacitors can explode merely because of a manufacturing defect at the vendor.  Something unable to detect before initial application of power.

Since I'm an Engineer and not a Lawyer or Safety Expert, I cannot advise you as to how to proceed other than to say you should take reasonable precautions to protect yourself and others from such dangers.  The extent of what those precautions are should be defined by your employer or research facility and should be aligned to comply with the laws of your local region.

Jim Noxon

Tantalum capacitors come in both wet (fluidic ectrolytic) and dry (solid) forms.  See http://en.wikipedia.org/wiki/Tantalum_capacitor for some additional descriptions.  Either form can burst due to reverse bias voltage application.

Just for reference, and purely for informative purposes, I can remember a time (long ago) when we would have whole boards of 14 to 20 pin DIP style chips populated each with one or more tantalum bypass capacitors next to each chip.  Since the equipment these boards worked in would get fairly warm, the capacitors would regularly degrade to a point where eventually one or more of them would provide a relatively low impedance path between the voltage rail and ground.  Since there were a couple hundred capacitors on these boards, it was extremely time consuming to desolder each one and then retry the board in the system to see if you found the failing one.  To speed up the process, we would remove all the chips from their sockets and apply about 25V in reverse polarity to the power rails on the board.  The weaker caps would quickly "Pop" and once you heard a pop you would remove the reverse voltage applied.  Then visually inspect the board for which cap blew and replace it.  It was a bit risky and probably weakened otherwise healty caps on the board but it got the equipment up and running much faster and saved clients significant amounts of repair cost even at the risk of reducing the life of other caps on the board.

At the time, I never concerned myself with the questions of safety you have brought up in this thread although I probably should have.  I was much younger then and didn't require wearing the glasses I currently wear for my failing vision so even this amount of protection was not utilized.  In fact, I can remember trying to look directly at suspected caps when doing this to see if I could watch one fail.  Sometimes, looking back I wonder how I've survived as long as I have.

Regardless, you are correct to be concerned about safety issues with board bring up and initial power on testing of manufactured boards.  I'm sorry I don't have any real answers for you beyond taking reasonable precautions.

Jim Noxon