'm usually in the position of trying to use basic statistics and common sense to combat fear-mongering about the dangers of terrorism, but I have to tell you that you are just plain wrong about the impracticality of binary liquid explosives.
I worked with concentrated peroxide for years at Armadillo Aerospace, and I can tell you with 100% certainty that you can reliably mix two clear liquids and get a high explosive that can be detonated by a sharp impact.
TATP is an unstable molecule that can slowly form in acetone / peroxide mixtures, even fairly dilute ones. Synthesizing and purifying that is probably something better done in a lab, but a simple solution of concentrated peroxide and acetone (or any other soluble fuel like methanol or ethanol) is also an explosive that is just as powerful. Many explosive, like gunpowder or ANFO, are not compounds at all, but simply mixtures of oxidizers and fuels.
You do need concentrated peroxide to do this, the common 3% peroxide from the drug store won’t cut it. We have first hand data points at a couple of these concentration levels, but there are some papers available that map out the sensitivity / concentration / mixture ratio diagrams with a lot more tests.
50% peroxide / fuel mixtures won’t detonate, even with a blasting cap. We used a very large amount of this safely as a rocket propellant, which does demonstrate that just mixing them is not hazardous.
60% peroxide / fuel mixtures can be detonated with a blasting cap. A blasting cap would probably be caught by explosive sniffing equipment, so it probably wouldn’t be useful for terrorists.
85% peroxide / fuel mixtures can be detonated by modest impacts. Shaking a plastic bottle of mixed solution with a few ball bearings in it would probably do it.
98% peroxide / fuel mixtures can be detonated by static electricity.
30% concentration peroxide is available in jug quantities from many places on the net. 50% peroxide is available in drums from many suppliers. We bought over ten tons of it while we were doing mixed-monoprop propellant work a couple years ago. 70% peroxide is what most manufacturers produce, but it only gets delivered by tank car to qualified receiving stations. The only remaining domestic supplier of 90 / 98% peroxide essentially only sells to government contractors.
This does mean that you can’t just pick up the chemicals you need, but don’t overstate the difficulty of doing the concentration. We did decide that it wasn’t worth our while to do large scale (tens of tons) concentration ourselves, but here we are talking about just making enough to fill a couple water bottles. We concentrated a few gallons ourselves early on with vacuum distillation, and there are a couple people you can find on the net that will sell you a ready-to-use concentrator system for modest quantities. For the water bottle quantities, all you have to do is leave some 30% peroxide in an open crock pot under modest heat for a few days, and you are left with a much smaller quantity of 85% peroxide. A member of the Sacramento L5 society wrote a nice report on this process.
There is no immediate reaction when mixing these fuels into even 98% peroxide. Left on a shelf for a long time, or in the presence of various chemicals that can catalyze the reaction, you will get some of the organic peroxide molecules that chemists usually think about when you mention peroxide explosives, but this isn’t at all important for just making a bomb.
I can say this with certainty, because we have actually done it. When we began working with concentrated peroxide as a rocket propellant, we did many tests to sort out the mythical dangers from the real dangers. One of these tests was the explosiveness of peroxide / fuel mixtures. We used methanol with 98% peroxide, but not at an optimal mixture ratio. It would do just fine to blow a hole in the side of an aircraft. No doubt about it.
I wouldn’t advocate any additional regulation or security measures based on this, but promulgating a new myth that binary explosives aren’t a valid danger isn’t helping anything.
John Carmack
