The Case of the Burnt Bowls

Setting: A well lit office and laboratory in Burlington, Ontario. Jamie answers the telephone.

Jamie (ISI): Hi Rick, how are you doing and what's up?

Rick (Forensic Engineer): An odd fire down near St. Marys, flash kitchen fire with some stuff on the stove. I'll send Sharom back with some stuff, he'll fill you in.

Jamie: Sounds good, I'll be in this afternoon and we can likely get right on it.

Later that afternoon. Sharom enters the laboratory with the evidence.

Jamie: Hi, Sharom. Rick called, what do you have?

Sharom: Well, these bowls were on the stove and that appears to be the origin of the flash fire, although it's hard to tell. The place wasn't badly damaged but these bowls are pretty burned up inside. I guess we need to know what the guy was doing on the stove before the insurance can be paid. Any ideas?

Jamie: It looks like the guy was cooking something. It's pretty black. I suspect there isn't much chance of finding residual liquid accelerant, although the residue itself might tell us something. We'll take a look at it in the lab and I'll give Rick a call tomorrow with our ideas.

This was the start of an interesting case back in 1993. The names (except the author's) and places have been changed as criminal proceedings may still be active. Amongst other scientific and consulting services at ISI, we have been supporting the forensic engineering, industrial, legal, and insurance industries with forensic analysis and problem solving for 10 years now and, in that time, have participated in many routine and non-routine cases. This is one of the less routine problems that required a solution. On with the case!

The evidence was a pair of standard stainless steel kitchen mixing bowls with burn marks and a small amount of black, burnt greenish oily residue in the smaller bowl. We inspected the bowls and residue and our senior chemist performed a few quick spot tests and solubility checks. We postulated that the residue contained an oily component (ie. solvent soluble) and was likely a plant-based product but could be petroleum based. It looked like the bowls were used like a double boiler, with water in the larger one and the original oily substance in the smaller one. This type of rig would be used for an extraction or to heat something in a controlled fashion to 100 degrees celsius.

I called Rick back and we discussed the details. It sounded like the guy whose house burned had a pot farm in the basement, with lights and a few dozen marijuana plants. He was known in the area. We agreed we would test the residue for materials consistent with cannabis extraction.

The odour had made us suspicious before. We checked the literature and determined illicit cannabis extractions are generally done with a light organic solvent - hexanes, light petroleum distillates, or alcohol. This was consistent with what we had observed and also with the flash fire, often caused by the ignition of vaporized light solvents. Our experience with botanical extractions told us that a dichloromethane extraction and GC/MS analysis of the residue would show most of the common plant oils such as phytol and the literature indicated that the compounds diagnostic with cannabis (tetrahydrocannabinol and cannabicyclol) should also be apparent.

The residue was removed from the bowl and a portion extracted with dichloromethane, dried, and run on the GC/MS (30 m non-polar column) under open scan conditions. The resulting chromatogram and computer spectral match are shown on the following page. As can be seen, the largest peak in the extract was phytol (1550 scans) which is characteristic of any 'green' plant extract (ie. leaves, stems, flowers but not necessarily fruit or roots). The second largest peak (1990 scans) was identified by the GC-MS computer with a 99.0% match as delta 9-tetrahydrocannabinol, the main psychoactive ingredient of cannabis. There was not much doubt about the identity of the material. The compound cannabicyclol was also detected at 1840 scans.

Page 1 2